BTW- if you’ve subscribed to this newsletter for insights about audience attention, I now do more writing about that on our Storythings newsletter Attention Matters. I’m going to use this newsletter to post chapters and insights from the book, so if you love really niche research into the history of attention metrics, stay here! If not, you might want to unsubscribe from this newsletter and join Attention Matters instead:

Attention Matters

Weekly research on audience behaviour - get the attention your content deserves.

By Hugh Garry

Introducing Mr X

It’s 8pm, February 16^th, 1964, and something momentous is about to happen. In living rooms across America, a newly familiar ritual is being carried out, as dinner is cleared away and families gather around the television. In the last decade, the number of TV sets in America has doubled, and fifty-one million of them are being switched on, vacuum tubes warming up, fuzzy black and white pictures coming into focus. Like almost half of America - and pretty much every teenager - you’re in your living room, tuning the TV to CBS for The Ed Sullivan Show.

The show opens with the sights and sounds of a horse race, crowds cheering as the horses cross the line, the camera cutting to a long shot of the finish as the show titles are superimposed over the top.

“Good Evening Ladies and Gentleman! Tonight, live from Miami Beach, the Ed Sullivan Show! Tonight the show is brought to you by Lipton Tea, your change of pace drink. Your change of pace and flavour of refreshment. Brisk Lipton Tea.”

This is not what you’re waiting expectantly to watch. You’re not sitting there - butterflies in your stomach, curled in a knot on the sofa or fidgeting on the floor - waiting to hear about Lipton Tea.

“And now from the stage of the Deauville Hotel, here he is, Ed Sullivan!”

The picture cuts to a long shot over the heads of a restless looking crowd. In the distance you can see the stage, thick drape curtains, various figures moving cameras and equipment. Cut to Ed Sullivan in a sharp suit and pocket square, slicked-back hair in a widow’s peak, hunched shoulders and his long, hang-dog expression. “Thank you very very much, thank you” he says, arms coming out to still the crowd’s applause.

“Well you know, it’s so very nice to be here, thank you Ralph Renick for that nice intro. And now, this has happened again. Last Sunday, on our show in New York, The Beatles played to the greatest TV audience that’s ever been assembled in the history of American TV. Now tonight - here in Miami Beach - again The Beatles face our record-busting audience. Now before I bring on The Beatles, lets… quiet!”

Sullivan holds out his hands again. He was stiff and unbending for this first monologue, arms tight by his side, shoulders bouncing up and down like a marionette. Now, as he barks ‘quiet!’ he holds his arms out wider and smiles, the crowd takes the cue and laughs nervously, and he visibly relaxes.

“Let me tell you the weather here is sensational, even our Californian friend George Fenneman agrees with me”

He draws out the last syllable in ‘agrees’ and slides off camera to the left as George Fenneman steps in to audience applause. He looks exactly the same as Sullivan - sharp suit, slicked-back hair, widow’s peak.

“I certainly do Ed. Yes, it’s always pleasant working here in Florida, the people are really great. Now, for instance, let me show what happened today, out at the pool.”

The picture fades and opens up on an establishing shot of the Deauville Hotel, then cuts to a close up of George Fenneman by the pool, a box of Lipton Tea placed on the table in front of him.

The show has been on two minutes already, and we’re still watching Fenneman by the pool, still in a sharp suit, but seersucker instead of dark wool. It seems to go on for ages. Then - finally! - it cuts back to Fenneman on stage. After a last plug and show of the Lipton Tea box, he hands over: “And now, Ed Sullivan.”

You can hear the crowd again, restless like you, applauding with the odd scream or squeal in the background. There are hums and noises coming from behind the curtains, sounds of amps and instruments being plugged in. Ed Sullivan comes in again from the left, shakes Fenneman’s hand and looks at the camera. He doesn’t look hunched up now, but is smiling, full of energy, his body swaying as he talks.

“Ladies and Gentleman”

A pause, a long look down the camera.

“Here are four of the nicest youngsters we’ve ever had on our stage, THE BEATLES - bring em on now!”

As he says their name, screams erupt from the crowd, and Sullivan throws his right arm across his body, spinning around like a pitcher delivering a fastball, all his nervous energy exploding until he ends up with his back to the camera, eyes looking back to where he’d just entered, as if he has to shield himself from what is about to happen.

Finally, three long, tortuous minutes into the show - there they are. Paul, George and John burst forward to take their mikes, George hitting a bass note on his guitar as the curtains part, Ringo smiling from behind the drum kit. Paul looks back at Ringo, and shouts ’1,2!’ There’s a rumble of drums, clanging guitar chords and three voices in close harmony:

“She Loves You Yeah, Yeah, Yeah!”

Live, from Miami, into over fifty million living rooms across America, come The Beatles. The cameras in the Deauville Hotel beam out their signals to New York, Colorado, Michigan, Kentucky; picked up by aerials fixed to suburban rooftops, or wires propped on bookcases. Fifty million vacuum tubes, framed in ornate wooden surrounds, trace the images of four young men from Liverpool onto domed glass screens.

Their shiny grey suits are a contrast to the sober uniforms of Ed Sullivan and George Fenneman; tighter fitting, with contrasting dark collars. Despite it all appearing in black and white, they seem like a flash of colour, all shaking mop hair and jerking guitars. Sometimes, in close-ups, the primitive technology can’t cope with the contrast between them and the stage behind, and they are ringed in a halo of dark shadow, as if they were cut out from their surroundings.

During the third song - All My Loving - Paul sings the line “I’ll pretend that I’m kissing, the lips I am missing, and hope that my dreams will come true” and as the shot closes in on him, he looks straight down the lens, through the cameras, across the air, down the aerial, into the vacuum tube, onto the domed glass - and straight at you.

The Beatles have been dominating radio stations since late 1963, forcing Capitol to release their album earlier than planned,  but what America wanted was not just to hear them, but to see them.

And now they’re here, in your living room, as they play to a packed ballroom in Miami. And you’re part of it, part of the biggest TV audience America has ever seen, over 70 million people in 50 million homes, all watching at the same time. Only television could create the illusion that The Beatles were in your living room, and Paul McCartney was looking straight into your eyes.

A few hours later, back in New York, a man twice the age of The Beatles is about to appear on another CBS TV show. He too is dressed in a grey suit, but not the same tight-fitting cut as The Beatles, and without their dandy contrast collars. Instead of a mop top haircut, he is completely bald, with small round wireframe glasses.

The show is filmed in CBS Studio 52, just around the corner from the more impressive Studio 50 where The Ed Sullivan Show is usually recorded. Just seven days earlier, Studio 50 had been the location for The Beatles’ first US TV appearance.

Studio 52 is more modest, a converted musical theatre adapted for TV and radio game shows. At 10pm, a light jazz theme tune plays over the cartoon  intro to What’s My Line? 

The four hosts - Dorothy Kilgallen, comedian Buddy Hackett, talk show host Arlene Francis and Bennet Cerf, the founder of publisher Randon House - take it in turns to introduce each other and take their seats. Cerf is last, and finishes by introducing their host.

“And now it’s my pleasure to introduce our panel moderator, who causes riots wherever he goes - and remember all you beautiful little girls in the audience you promised me you wouldn’t go into hysterics when I introduced him - John Charles Daly!”

The joke is a callback to The Beatles TV appearance around the corner in Studio 50 a week ago, and then again, just a few hours earlier that night, in Miami. The audience laugh, and then dutifully scream as Cerf waves his hands in mock astonishment. The laughter continues as Daly comes on stage and takes his seat behind the desk across from the panel. Like the other two men, he’s in a sober black suit with bow tie. 

The contrast between the five people now on stage in Studio 52 and the four men who just appeared in Miami could not be greater. The shock and impact of The Beatles’ second US TV appearance has caused ripples through the rest of television, so even a show as tightly formatted as What’s My Line? - now in it’s fourteenth year - has to find a way to acknowledge it. But Daly soon settles the show into a familiar routine, gently chastising Cerf for his intro, and sharing a small bit of repartee with Hackett about how he’s been helping him with his golf game. And with everything back to normal, we’re ready for our first guest.

“And now to meet our first contestant, would you enter and sign in please?”

The camera cuts to a tight close up of a blackboard. There’s the sound of a few soft footsteps off camera, then an arm appears from the left holding a piece of chalk. The arm draws the words ‘MR X’ on the blackboard, drawing a circular full stop after ‘MR’ with a flourish. Daly announces ‘Mr X’ and the camera cuts to the back of the guest, a balding man in a grey suit and round glasses. He shakes hands with Daly and takes his seat to the left.

“Now for reasons that will be clear later, we are presenting our guest as Mr X, but let the folks in the theatre and the folks at home know exactly what your line is.”

The camera closes in on the mystery guest as his job appears on the screen overlaid on him. The audience applauds politely and there’s a murmur of low conversation.

“Alright panel, we can tell you Mr X is self-employed and deals with a service and we’ll begin the general questioning with, uh, Arlene Francis.”

“Mr X, are you Mr X because your name has appeared in the paper, in the last, uh, month?”

Mr X looks to Daly who purses his mouth and nods repeatedly. He turns to Francis and responds. “Yes.”

“Is it a name, probably, well known to the panel?”

Mr X nods gently. “Yes.”

“Do you have anything to do, in any way, with the arts?”

Mr X thinks for a second, then turns quickly to look at Daly. The camera pans out to include them both.

“Well I would think we would have to agree that you have something to do, in any way, with the arts, yes?” They both look back to Francis.

“Is it one of the graphic arts?”

Mr X looks back at Daly, who purses his lips.

“Mmm, no, I would say not. That’s one down and nine to go. Mr Cerf.” 

Daly flips over a wooden board at the front of his desk, changing it from a question mark to ‘$5’. Mr X shakes his head in agreement. Bennett Cerf asks his first question.

“Mr X, has the work that you have been doing, and which has brought you the fame which has brought you I’m sure to this programme, go anything to do with the government?

Mr X shakes his head “No.” Daly flips over another board that says ‘$10’ “Two down and eight to go - Miss Kilgallen.”

“Ah Mr X can you conduct your work wearing an ordinary business suit?”

“Yes.”

“Do you work in the daytime?”

“Yes.”

“Do you usually work indoors?”

“Yes.”

“Uh, do you have anything to do with one of the performing arts, in some sense?”

Mr X looks at Daly again, who repeats the question, “Do you have anything to do with one of the performing arts in some sense? I’d say so, yes.” Mr X nods in agreement and says “Yes”. They both look back towards Francis.

“Would it be, uh, the theatre or the opera?”

Mr X says “No” and Daly flips over another board that says ‘$15’. “That makes it three down and seven to go - Mr Hackett.” The camera moves to Buddy Hackett, looking askance at Mr X through squinted eyes. As he asks his question his jaw chews on the words as if they were bubble gum.

“Mr X, I take that X, uh - this answer if I’m phrasing it right should be ‘yes’ - the X is not half of a name like ‘secret agent X9’ - is that right?”

The crowd erupt in laughter and the camera cuts to Mr X and Daly laughing and nodding.

“Mr X, when you go out of doors, do you wear a hat?” The laughter increases as the camera cuts to Mr X, his bald pate gleaming in the studio lights. 

“Ever since Buddy adopted that Beatle wig of his there, you know…” Daly responds, and Hackett tugs his thick black forelock - “It’s real, you know! Mr X, would you like to have hair? That’s another question!” Mr X and Daly continue laughing as Hackett asks his next question.

“Do you have anything to do, er, with the new Lincoln Center?”

Another small shake of the head from Mr X - “No.” Daly flips another board showing $20. “Four down and six to go - Miss Francis.”

“Since you have ruled out the graphic arts, may I assume you have nothing to do with a drawing board, whatsoever?”

“Yes, that is correct.”

“Ah-huh. However, what you do is creative, is that correct?”

There’s a small nervous laugh in the crowd. Mr X looks at Daly, who seems equally perplexed. Daly says “We’ll have to have a small conference.” and they both lean in for a private conversation, Daly’s hands hiding their mouths from the panel. “We say no. That’s five down and five to go. We’re assuming here that you are using the word ‘creative’ in the artistic sense, and we would have to agree here that the answer should be no. Mr Cerf.”

The camera cuts to a confused looking Arlene Francis, her mouth dropped open, then pans to Bennett Cerf. 

“Mr X, we did elicit from you the fact that you are somehow connected, in some way, with the performing arts, but not the theatre and not motion pictures, is that right? Might you have anything to do with television or radio?”

“Yes.”

“Would it be television?”

“Yes.”

“Uh, but you are not the representative of any government or state commission, is that correct?”

“That is correct.”

“You are self-employed. Have you got anything to do with the ratings of television channels?”

The camera cuts to Mr X, who nods and says “Yes”, and Daly, who is suppressing a grin.

“Are you Mr Nielsen?”

Daly exclaims “Yes!” and bounces up in his chair, applauding, the audience quickly following, the panel now grinning in acknowledgement.

“May I formally present Mr A.C. Nielsen Jr., who is the President of the Nielsen Service.”

Arlene Francis jokes “We have been led to believe that is creative in a way!” and everyone laughs as Daly replies “Actually, that’s why I came back and said ‘in an artistic sense’”

A nervous excitement takes over the panel as they realise the influence and power Mr X has over their television careers. “Be careful what you say you fools!” Cerf laughs nervously. Buddy Hackett rests his hand lightly on his fingers, and looks seriously at Mr X through squinted eyes.

“On behalf of Max Liebman, I’d like to say a word to you about a show I once had called ‘Stanley’…”

Hackett had been the star of Stanley, a sitcom produced by Max Liebman, for just one season in 1956-57. It was cancelled after consistently poor ratings. Mr X replies “Well, this particular show I hope gets a very high rating.” The panel and audience laugh, and Daly gestures to Dorothy Kigallen who straightens her back and speaks deliberately, as if reading from a card - “I just want to say on behalf of What’s My Line, we love Mr Nielsen.”

The panellists and host are now flattering Mr X as if he was minor royalty, Daly tripping over his words as he quickly agrees. 

“We do. We love the public who makes it possible for Mr Nielsen to make it possible for us to be lovely to Mr Nielsen into the public, or something like that, how’s that for everything? I must say, it’s been a lot of fun, I, er, I didn’t think they’d catch it, but they did in fact when they got into the arts I thought anybody got into television I was afraid we’d get hooked, thanks very much, nice to have you with us.”

Mr X leans over and shakes hands with Daly, then rises and walks to the panel to shake their hands - Cerf and Hackett standing to meet him, Kilgallen and Francis remaining seated - before leaving the set to the left.

After the audience applause finishes, Daly flips the boards back to the question mark.

“Allright, now to meet our next challenger.”

For just over 8 minutes, on the evening of Sunday, Feb 16^th 1964, Arthur C Nielsen Jr. stepped out in front of the audience, briefly tasting the glamour and fame of a television appearance. Although his face was anonymous, his name was well known, especially to the talent on screen, who’s careers depended on his data.

In 1964, the year The Beatles and Arthur C Nielsen Jr appeared on TV just a few hours apart, the amount spent on TV advertising in America had reached $2.28bn, second only to newspaper advertising. By 1973, it would almost double again to $4.5bn

But these billion-dollar industries were based on an incredible tiny amount of data. A 1965 CBS documentary called The Ratings Game revealed that Nielsen’s sampled data relied on just 1,130 ‘Nielsen families’ - households who agreed to have the Nielsen measurement technology installed in their homes so that their viewing patterns could be recorded and analysed.

Each one of those thousand or so Nielsen families had an economic and creative power that would have been almost impossible for them to imagine. Every decision they made to switch on the television or change the channel directly affected the commissioning decisions of the emerging TV broadcasters, the budgets of advertising campaigns for Madison Avenue advertising agencies, the rise and fall of share prices on the New York Stock Exchange, and, as Buddy Hackett had sourly noted, the careers of a new generation of TV talent. Each Nielsen family represented about 45,500 other families - about the size of a small town - and their decisions influenced around £2m of TV advertising spend. Without even knowing it, they were kingmakers, wielding power like the gods of Ancient Greece, but with buttons, dials and  remote controls instead of bolts of lightning.

So how did television - the medium that more than any other defined the twentieth century - come to rely on such a small number of people for their ratings? And how did the Nielsens - father and son - end up with such a monopoly over the industry that they became a household name?

The Nielsens’ empire was built on solving a simple problem - how can you measure what people are doing when you can’t see or hear them? The technology revolutions of the 20^th century created powerful communication networks - first radio and  television, then later the internet - but as the power of these networks was their ability to cover great distances, they created a new problem. If you were providing a service over these networks, how could you measure your audience when they were hundreds of miles away, sitting in their living rooms?

Back in 2003, when I worked at the BBC, a friend suggested writing scenarios about the future of digital media. This was two decades ago, right at the birth of Web 2.0, a time of huge optimism about digital change, and I was balanced between the old and new worlds, working at the new media dept of the world’s oldest and most respected broadcaster.

I decided to write my scenarios in the form of a set of questions for a Media Studies exam paper in the, then unimaginable, future of 2018. Every now and then I go back to look at it again, and there are a few things in the questions that I think I got kind of right (particularly the ones that talk about identity, privacy and truth). But there’s one question in particular that I keep coming back to, because the future I was imagining through it has taken far longer to arrive than the 15 years I suggested back in 2003. Here’s the question:

3) What was ‘Scheduling’? Discuss giving examples from the major linear media of the late 20th Century.

I remember being really pleased with the use of the past tense in the question, imagining that TV schedules would have receded so far into the past by 2018 that teenagers sitting exam papers would need to write about them as historical objects.

Clearly, I was very wrong, and twenty years on scheduled media is still holding on by its fingertips, the curve of it’s decline over audience attention moving inexorably, but slowly, downwards.

But a couple of things that have happened recently make me feel that we might only see linear scheduled channels around for, maybe, another decade. So my initial prediction of 15 years was perhaps only half right.

The 2023 OFCOM Media Nations report shows that even the most loyal schedule-viewing demographic - over 65s - are now watching less live TV and more streaming. Overall tv/video viewing was down from it’s covid-era peaks of 2020/21, but linear viewing is falling as streaming and other IP viewing takes a bigger share. Live TV has been less than 50% of all video viewing for the last few years, although the traditional PSBs (public service broadcasters) still have 60% of total viewing thanks to their IP platforms:

When you break down this viewing by age cohorts, you can see just how much live scheduled TV viewing is held up by older audiences, as it just a tiny fraction of younger audience’s video viewing:

But it’s not these audience trends that are the potential tipping point I’ve been looking for over the last two decades. The real signs that we’re seeing the end of traditional scheduled TV come not from demographics, but the business models of some of the biggest traditional media companies in the US and UK.

First of all, US cable company Charter released a remarkable investor deck about their spat with Disney on the fees for including ESPN and other linear video channels in their cable bundles. The deck illustrated how the current video ecosystem is broken, as linear video owners tried to keep carriage charges to cable companies artificially high, despite the traditional cable companies losing 25m customers to ‘cord cutting’ in the last 5 years. Its an astonishingly frank and honest deck, laying out how the current business models of linear TV are on a death spiral that could take down both the cable companies and the linear channel owners. That honesty seems to have worked, as the two companies announced a deal shortly afterwards for a hybrid model, with Charter bundling Disney+, their DTC VOD app, along with the linear channels.

Then just yesterday a consortium of the UK Public Service Broadcasters (PSBs) announced a new service called Freely that will deliver live broadcast and VOD via a single app. The project will be run by a new company - Everyone TV - jointly owned by the BBC, ITV, Channel 4 and Channel 5. All of the PSBs already have their own VOD products, most of which let you stream live TV channels alongside VOD content, so the interesting signal here is why they felt the need to work together on a single app.

The strategy is similar to two previous collaborative projects - Freeview and FreeSat - that were strategic reactions to previous digital change in broadcast TV. In both those cases, the PSBs worked together to provide free alternatives to pay TV channels launched by the then Murdoch-owned Sky TV, and cable networks run by Virgin and other providers.

But the rise of smart TVs and VOD apps from Netflix, Amazon, Disney and others created a much more substantial strategic risk, one in which viewers navigated not through a schedule, but through the different interfaces of the apps they subscribed to. As OFCOM’s Media Nations reports have shown, audiences have quickly moved their TV habits towards VOD viewing, especially in the COVID years of 2020 and 2021, when viewers wanted to escape the doom and gloom of news on scheduled channels to the bright fictional worlds of VOD streamers.

Freely is the PSBs strategic response to this, an attempt to maintain their ‘due prominence’ in a world where regulation cannot mandate audience attention in the way it did in the scheduled TV era. It points to a really interesting insight into what might be the best strategic option for a post-schedule TV world. One of the challenges of the move to VOD has been the ‘unbundling’ of TV into multiple apps and subscriptions. This unbundling particularly affects live sports, one of the few areas of traditional TV schedules that can still attract large audiences, but the PSBs long ago lost the most valuable sporting rights to pay TV competitors. But even outside sport, there are signs that audiences might be reaching peak subscriptions, with newer channels like Lionsgate and Paramount struggling to drive take-up.

This is where a new ‘rebundling’ strategy could work. Both Disney and Charter and the Freely consortium are betting that the final transition from the traditional schedule to a fully internet-delivered broadcast world will not be smooth, and audiences will gravitate to bundles that make this transition less complex. This will be particularly true for older audiences, who might worry more about setting up lots of small monthly payments that are difficult to cancel, learning from their experiences with mobile phone and broadband providers.

Having worked at both Channel 4 and the BBC, I know how hard it is to get multi-stakeholder projects like Freely agreed. In fact, an earlier attempt to create a joint VOD service from the UK PSBs - Project Kangaroo - was killed by the Competition Commission in 2009, clearing the ground for the US-owned VOD platforms to come in and dominate a decade later. The fact that Freely is happening now, and that it appeared fully-formed with little gossip or fanfare, shows that the PSBs know this is their last chance to maintain prominence with audiences as they abandon the schedule.

If I was to update my 2003 scenario, I think I’d kick the date of the exam paper down the road a bit more, maybe to 2033, doubling my initial 15 year prediction to 30 years. And I’d add a bit of optimism, as my core belief in the social good of public sector broadcasting means I really want to see Freely succeed in helping PSBs transition from the era of the schedule to the era of the bundle. So here’s my updated version:

3) What was ‘Scheduling’? Discuss giving examples of how unbundling and rebundling changed the landscape of the TV and Video industries in early 21st century.

Come back in 2043 and we’ll see if I was right.

Before the web, Tim Berners-Lee wrote an application called ENQUIRE. It was based around cards that show the links between information in a database, allowing the user to move between the relational links as they research a subject.

The inspiration, and name, came from a book Berners-Lee had been given as a child by his parents:

“When I first began tinkering with a software program that eventually gave rise to the idea of the World Wide Web, I named it Enquire, short for Enquire Within upon Everything, a musty old book of Victorian advice I noticed as a child in my parents’ house outside London. With its title of suggestive magic, the book served as a portal to a world of information, everything from how to remove clothing stains to tips on investing money.”

ENQUIRE was limited to linking between specific databases of knowledge, so Berners-Lee abandoned it and moved on to develop a network that would allow linking across the boundaries of specific databases - the World Wide Web.

When we develop a new technology, the names and metaphors we use to describe them are critical. ENQUIRE was influenced by a book that claimed to be an oracle - to have the answer for every question, to promise the ‘suggestive magic’ that so excited the young Berners-Lee. But that promise was ultimately unfulfilled. Berners-Lee recognised that what he needed to build was not an oracle but a map - a web of paths and connections linking information and guiding users as they research, not an all-knowing entity that gives us a single answer.

The current frenzied debate about Large Language Models and Artificial Intelligence is partly because tools like ChatGPT look, on the surface at least, like oracles. We type in questions, enquiring within upon anything, and get something that looks remarkably like an answer. But what if these new technologies aren’t oracles, but a new kind of map?

The metaphors we use to describe LLMs and AI now are critically important. If we anthropomorphise them, and think of them as conscious, it will be almost impossible to not consider them as oracles. We will worry about them replacing humans in various tasks, worry about whether they are telling us the truth, and look for ways to regulate the ‘suggestive magic’ they seem to posses.

We’re making this assumption because the most uncanny version of these technologies are using conversational interfaces as their mode of interaction. There was a spike of interest in conversational interfaces a few years ago, as voice technologies like Alexa and Siri entered our lives. The difference between screen-based search and voice based search is the difference between a map and an oracle - on screens we’re given a list of search results to navigate, whilst in voice based search we’re not read a list of potential links - we’re given an answer.

When you are trying to build an oracle, you have an incredible responsibility to ensure that this answer is right. I was at a workshop with a UK Health Charity a few years ago in which someone from the NHS said that they had been approached by one of the big tech companies, asking if they could use their database to provide answers for their voice interface. If the oracle gives you the wrong answer when you’re looking for a movie recommendation, that’s annoying. When you’re asking for medical advice, it’s potentially harmful.

Most search engines generate their results using a map called a ‘knowledge graph’. This is a map of connections between objects that not only knows how these objects are connected to each other, but also knows what those connections mean. In 2007, Danny Hillis built Freebase, a knowledge graph that scraped data from the emerging networks of web 2.0, mapping the different types of connections between them. By encoding these types of connection within the data, it meant that search engines could do some of the work that users had previously had to do in deciding what result was the best answer. Hillis described it as a “digital almanac”, and web 2.0 entrepreneur Tim O’Reilly called it “like a system for building the synapses for the global brain.”

Metaweb, the company that built Freebase, was acquired by Google in 2010. As part of the deal, Hillis required Google to keep the knowledge graph public via an API for at least 5 years. In 2015, Google shut down the Freebase API and launched their own Knowledge Base that now runs the ‘knowledge panel’ you see on Google search results.

A couple of years ago, my company, Storythings, did a bit of comms work with MIT Knowledge Futures Group, where Danny Hillis and SJ Klein are working on a new version of Freebase called The Underlay. This updated version is based not just on helping machines understand the links between things, and what those links mean, but also the context of those links - who created them, and whether the links are reliable. The Underlay is designed to enable the dynamic presentation of these concepts and their relationship, a dynamic map of the ever changing relationships between facts. It’s not an oracle of knowledge, but a map of how we produce knowledge:

“The Underlay will gather knowledge currently used to produce publications, databases, and dynamically generated displays. It will make each associated assertion available in a machine-readable form that can be dynamically searched, vetted, and combined, based on its provenance. By connecting multiple sources together, each asserted claim can be analyzed for relevance and veracity, recombined and re-presented for different purposes. The Underlay will allow intelligent software to help people find what is relevant and judge what is true.”

The discussions around LLMs at the moment are mostly framing them as oracles. The hallucinatory conversations journalists are having with ChatGPT make it very hard to not anthropomorphise the technology, to see them as oracles, not maps.

But the best metaphor I’ve seen for LLMs, the one that has really helped me understand what they actually do, is a lot more like a map than an oracle. In his newsletter Humane Ingenuity, Dan Cohen, the Dean of NWU and academic on digital media and humanities, describes the output of an LLM as not a single truth, but a path taken across a topological map:

“a simple LLM has the same issue a pool table has: the ball will always follow the same path across the surface, in a predictable route, given its initial direction, thrust, and spin. Without additional interventions, an LLM will select the most common word that follows the prior word, based on its predetermined internal calculus. This is, of course, a recipe for unvaried familiarity, as the angle of the human prompt, like the pool cue, can overdetermine the flow that ensues.

To counteract this criticism and achieve some level of variation while maintaining comprehensibility, ChatGPT and other LLM-based tools turn up the "temperature," an internal variable, increasing it from 0, which produces perfect fidelity to the physics, i.e., always selecting the most likely next word, to something more like 0.8, which slightly weakens the gravitational pull in its textspace, so that less common words will be chosen more frequently. This, in turn, bends the overall path of words in new directions. The intentional warping of the topological surface via the temperature dial enables LLMs to spit out different texts based on the same prompt, effectively giving the snowball constant tugs in more random directions than the perfect slalom course determined by the iron laws of physics. Turn the temperature up further and even wilder things can happen.”

I love the metaphor of the pool table - writing a prompt for an LLM is like pulling back the pool cue before you take a shot. You get to decide the spin, angle and power, and then the ball that represents your prompt ricochets around the topological surface of the knowledge space within the LLM, finding the optimal path that ends up shaping it’s response.

Just like the Underlay, LLMs are dynamic, and can be made ‘weird’ by changing the parameters used to shape the topological space your prompt ricochets around. This is what made me realise that they are maps not oracles. The hallucinatory responses in ChatGPT are not answers - they are paths within a multidimensional space of knowledge, helping us map out those spaces and understand its hills and valleys.

The reason we don’t immediately recognise these tools as maps is partly because of the conversational interface, but also because the spaces within LLM are multi-dimensional. They’re not 2D or 3D spaces, but have 1000s of dimensions, and are incredibly hard for us to conceptualise as spaces. They create each new path anew, testing each step against the unimaginably complex probabilistic space of the corpus of knowledge they are trained on. They are maps of places that we can’t comprehend, and that is why we haven’t originally understood them as maps.

It’s also why ChatGPT stumbles on the kinds of truths that the Underlay is designed to resolve. If the topologies of the knowledge it is based on tends towards a certain path, that doesn’t automatically mean that the path is the correct one. The answer that it thinks is right is the one that looks like the most trodden path within the forest of knowledge it has access to.

This is why these new tools are best at mapping spaces that our previous tools haven’t been able to map. LLMs map the contours of massive corpuses of knowledge in a way that can reveal biases and structures in really interesting ways. In his New Yorker essay, sci-fi writer Ted Chiang described LLMs as a “blurry JPEG of the web”, in a way that was intended to be critical. But this looks at the problem from the wrong way round. LLMs optimal use is not to see things that we can already see clearly in a new way, but to bring into focus questions that we’ve never been able to ask before. It’s like the difference between the Hubble and the James Webb telescope - it’s best at bringing really fuzzy spaces into a bit more resolution.

The best illustration of this I’ve seen is Noah Brier’s BrXnd.ai project. Brier started with a simple toy called Collxbs that used ChatGPT and image AI DALLE2 to generate fictional images and text descriptions of fictional collaborations between brands. Brier was using the technologies to do what they do best - hallucinate imaginary answers that can only be produced by ricocheting prompts around the corpus of data the LLMs had about the brands.

But as he saw the outcome of prompts, he noticed that some brands had stronger presences in the fictional collabs than others. It was as if the LLM was intuiting something about the strength of those brands, about how much presence and power they had in the data the LLMs were trained on:

“Because it has a general sense of concepts from its gigantic corpus, it can group brands in meaningful ways without any other inputs. The image below is a two-dimensional analysis (I did a 3d one as well) of a bunch of brands that various friends asked me to include (hence the random assortment). What’s surprising about the image is how unsurprising it is. The model seems to have an intuitive understanding of categories when given nothing more than the brand’s name and some indicator that it’s a brand. (I do that so when I ask for Apple, for instance, I get the technology company instead of the fruit.) Since it seems basically correct, it also raises other interesting questions, like why has it grouped Western Union as the closest non-wine brand to Yellow Tail? Or what puts Google closer to consumer brands than Facebook?”

There has been decades of theories about how brands work - how the strength and recognition of a product’s brand has an influence on the purchasing decision of the audience. But the picture has always been fuzzy, relying on audience surveys and market research that might prove correlation, but not causation.

Brier is using ChatGPT to bring a fuzzy answer into a bit more focus, in the way that only LLMs can do. It’s a bit like doing audience research in a multidimensional space with millions of contributors all at once. Brier is mapping a space in a way that’s never been possible before.

Maybe this is what LLMs can do best - map knowledge spaces to reveal their biases and assumptions. One of the things I’ve loved about LLMs is the creativity they’ve opened up in development communities - people are playing and testing the boundaries of these new tools in incredibly creative ways.

I see this work not just as technological exploration, but as a kind of media literacy. By testing the boundaries of their abilities and weaknesses, we’re beginning to understand what kind of maps they are, and what kinds of things we can use them to understand. This feels really exciting to me, way more exciting than the idea that they are oracles that can replace human thought.

The internet has never been an oracle, and the answers we can find have always been, and will always be, fuzzy. We are, again, building a new kind of map, one that, like the World Wide Web and Freebase before it, will give us new ways to understand, produce, and share knowledge. The people building new interfaces for these tools need to understand that, and help us see them as maps, not oracles.

In my last post, I looked back to the beginning of 2007, and two events that I think marked the start of the 21st century. It now looks like we’re at the end of the first era of this century - the era in which public social media platforms were dominant.

Like any good future scenario, there are different drivers for this, some technological (TikTok’s algorithmic innovation) some financial (Apple’s privacy decisions decimating Meta’s advertising income), some social (younger audiences leaving Facebook and Twitter for newer platforms), and some wild cards (whatever the hell Musk thinks he’s doing with Twitter). Regardless of what’s driving it, it definitely feels like the end of the era that started back in 2007.

Despite working for the last two decades in roles that have involved thinking about innovation, I’m lousy at predicting the future. But back in 2007 I published something that tried to explain how social media spaces were mapping on to what we, as users, wanted from them. The ideas in that post have stayed in the back of my mind ever since, and now, as the platforms that have dominated the last decade and a half start their decline, I wanted to revisit it to see if there are any clues for what might emerge next.

2007 was the peak of the Cambrian explosion of web 2.0, and I wanted a way of explaining these new technologies that didn’t focus on specific sites, but instead looked for deeper patterns. The model I ended up with described six spaces that I felt were being created in social media, each with different expectations and behaviours. Here’s how I described them back then, in August 2007:

As part of my current job1, I’ve been trying to find a way of describing social media spaces in a way that can be shared by both traditional media indies and digital media agencies. The former understand genres and formats, whilst the latter understand platforms and networks. After a few weeks of gradually finding out what doesn’t work, I’ve ended up adopting a more user-centred model, based on the assumptions users have about what they can *do* in certain kinds of space, who they’ll be doing it with, and what kinds of behaviours are expected. I’ve been meaning to write this up for a while, so here they are – six different types of social spaces, based on behaviours and expectations, not platforms, genres or formats. Caveat – this a crude analysis, and the examples are not exclusive – there are lots of overlaps between these spaces; and they exist both online and offline:

Secret Spaces
Behaviours: Private, intimate communication, normally with only one or two others, often using private references, slang or code
Expectations: Absolute privacy and control over the communication between users, and no unauthorised communication from third parties (eg spam)
Examples: SMS, IM

Group Spaces
Behaviours: Reinforcing the identity of a self-defined group, and your position within the group, eg ‘stroking‘ behaviour to let the group share a sense of belonging, or mild competitiveness to signal hierarchies within the group (eg who has the most friends, posts, tags, etc)
Expectations: A shared reference point for the group – eg a band, football club, school, workplace, region, etc. Rules about approving membership of the group, and icons for the group to signal their membership (badges, profiles, etc)
Examples: Facebook, Myspace, Bebo, etc

Publishing Spaces
Behaviours: Creating your own content or showcasing your talents to an audience outside of your usual social group
Expectations: The ability to control the context and presentation of your creative content. Ways to receive feedback, comments and advice from other users.
Examples: Flickr, Youtube, Revver, etc

Performing Spaces
Behaviours: Playing a defined role within a game structure. Experimenting through simulation, rehearsal and teamwork to achieve a goal. Iterative exploration or repetition of activities in order to perfect their performance
Expectations: A clear set of rules that is understood by all players. Clear rewards for success or failure. The ability to test the boundaries of the game structure, or to perform extravagantly to show off your talents
Examples: MMORPGs, Sports, Drama

Participation Spaces
Behaviours: Co-ordination of lots of small individual acts to achieve a common goal. Shared belief in the goal, and advocacy to encourage participation by others.
Expectations: Rules or structures that help co-ordinate activity towards the goal. The ability to create micro-communities within larger participation groups – eg a group of friends going on a political march together, or a workplace group created to train for a marathon
Examples: Meetup, Threadless, CambrianHouse.com, MySociety

Watching Spaces
Behaviours: Passive viewing of a linear event as part of a large group. Organising a group to attend an event, and sharing experiences afterwards
Expectations: Spectacle, entertainment, a feeling of thrill or joy. A shared sense of occasion, or of being taking out of your everyday existence for the duration of the event. Mementos or relics of the event (eg programmes, tickets, recordings, photos, etc)
Examples: Television, Cinema, Sports, Theatre, etc

Looking back, it’s interesting to see the examples I used - I’m surprised I didn’t include Twitter, which had launched the year before. But in general, I think the model has held up pretty well as a way of describing what we want from social spaces online.

At the time, I thought that the biggest issue with digital social spaces was going to be transgression - that the technology would move our social activity across these definitions, in ways that surprised us or worked against our expectations. In our work with young people at the time, we saw a definite line after the first two spaces that marked the boundary between spaces that felt private in some way, and spaces that felt like they could expose them to ‘randoms’ (as the terminology of the time described them). We were working on projects that tried to get young people to collaborate in public, but they saw that line as something they were often too nervous to cross.

What I didn’t understand then was how much the need to generate advertising dollars would push us all over that line, and eventually make almost everything we do public by default. In 2010, Matt McKeon published a great visualisation of how Facebook had shifted its default privacy settings over the previous 5 years, with nearly everything being accessible not only to Facebook users, but the entire internet, by default. If this felt shocking in 2010, it would only get more shocking over the next decade.

This feels like the real story of the last 15 years - the big platforms needed advertising to fund their incredible growth, and so ran roughshod over the promises they’d made about user’s control and privacy when they joined their services. If things are in public, they’re easier to monetise, and so this was the direction social platforms took - away from secret and group spaces, over the boundary into spaces where users were more public, more passive, and more monetisable. The social and political cost of this trangression has been incalculable.

The last fifteen years has seen social networks drift to the extreme ends of my six spaces model. At one end we had private messaging apps like Signal, WhatsApp and iMessage, and at the other the big platforms like Facebook, Instagram and Twitter, which, although they contained private messaging functions, really wanted you to hang out in public.

The spaces in the middle, where we were publishing content or trying to participate to achieve group goals, took a different turn. The rise of the creator economy turned publishing content from a purely social activity into a monetisable side-hustle, and sites like Flickr lost out to platforms like Instagram and Youtube that pivoted away from creative expression towards creative entrepreneurialism. Participation spaces likewise pivoted from sites that merely tried to enable community - like MeetUp - to ones that were designed to monetise it, like Patreon and GoFundMe.

The space that we were most excited about in our work with young people in 2007 was the ‘performing’ spaces, where you could play a role that helped you test and rehearse your abilities and identities. This is most obvious now in massively mutliplayer online games like Fortnite, and yet is still, I feel, hugely misunderstand when people talk about the growth of social spaces in the last 15 years. Zuckerberg’s vision of the metaverse feels like it’s extrapolated from business software, not the vivid pixelated dreamspaces that Minecraft and Roblox have enabled. It misses the most important factor that makes ‘performing’ spaces work - a set of rules that give you something to bounce off, that makes it fun.

But even though I’m terrible at predicting the future, I am still an optimist. As the major platforms seem to be waning, I think we’ll see a new mini-cambrian explosion of platforms and services that are more diverse, more focused, and that hopefully have more respect for the boundaries we expect in our online social spaces.

I can see that already in my own behaviour. Nowadays I socialise a lot more in private group spaces enabled by Slack and Discord than I do on public social media. And the newsletter boom supported by Substack and others has created new forms of community that feel closer to the blogging revolutions of the early 2000s.

The optimism of the web 2.0 era was engulfed by the tsunami of money that emerged from monetising our public social behaviour. As we start to pick ourselves up from the wreckage that caused, I’d like to regain some of that optimism again. I’d like to think the services that are emerging now will be more tightly aligned to the kinds of expectations and behaviours I described in the ‘six spaces’ model. I’d like to think they will have viable business models that are based around people paying for the things they value, not being turned into products for advertising.

Most of all, I’d like there to be a healthy ecosystem of smaller, more focused products, and not a handful of giant platforms dominating the public sphere. But as I said, I’m terrible at predicting the future.

[note - this is adapted from a draft of a chapter I wrote for a book on the history of attention metrics. It’s very unlikely to be part of that book now, but it feels appropriate, as we move to the end of 2022, to publish it somewhere. BTW - it’s really hard to find a video, transcript or detailed account of the Facebook launch event I describe here, even though it was only 15 years ago. I eventually tracked it down on an obscure Chinese social media site.]

History never quite aligns with the maths we use to divide it. When we flip from one decade, century or millenia to the next, we aren’t suddenly in a new world. The historian Eric Hobsbawm has argued that the 20th Century was ‘short’, a period of modernity that started with the gunfire of WW1 in 1914 and ended with the fall of the Berlin Wall in 1991.

As we’re nearly a quarter of the way through the 21st Century, could we say when our current era really started? I’d argue that it didn’t start with a single political event like 9/11, but with two events that have shaped how we see the world - our culture and politics - in ways that will take at least another quarter century to resolve. These events happened in the same city, barely a few months apart in 2007, and together they have had a profound influence on society, much more than was possible to understand at the time.

On Jan 9th 2007, Apple launched the first iPhone, and a few months later, on May 24th, Facebook launched the Facebook Platform, a tool that would let developers embed the Facebook network across the web. These two products - a mobile phone with sensors and cameras that capture a wealth of data about our lives, and a social network that can gather our actions and opinions from across the web - have, over the last 15 years, fundamentally reshaped much of our world in a way that makes those moments feel, more than anything else, like the start of a new century.

The first event is now legendary. On January 9th, at the Moscone Center, the biggest convention centre in San Francisco , Steve Jobs was giving his keynote at Apple’s annual Macworld event. Dressed in his trademark black polo neck, jeans and white sneakers, he walked across the stage looking at his feet, lost in thought.

“This is a day I’ve been looking forward to for two and half years.”

As he paused again, pacing across the stage, fiddling with the presentation clicker in his hands, a smattering of applause and a few whoops came from the audience. He stopped, looked up at the audience, and addressed them directly.

“Every once in a while, a revolutionary product comes that changes everything.”

There had been months of feverish speculation that Apple was about to launch a new phone, and Job’s presentation was a masterpiece of storytelling. But Jobs wanted to tease the audience first. He announced Apple was launching three new products - a widescreen iPod with touch controls, a revolutionary mobile phone, and a breakthrough internet communicator. Then he kept repeating these three products in a kind of mantra, as their logos spin on the screen behind him - “an iPod, a phone, and an internet communicator” - and the crowd starts laughing, anticipating the punchline.

“An iPod, a phone…. Are you getting it? These are not three separate devices. This is one device. And we are calling it - iPhone.”

It’s interesting, looking back at this keynote 15 years later, to see the audience’s reaction to the ‘three new products’ as Jobs announces them. The widescreen iPod gets applause and cheers, but the ‘revolutionary mobile phone’ really turns the audience wild - shouts of ‘yeah!’ ring out, and the faces of the attendees look almost in awe. The third product gets a more muted reaction, and Jobs seems to fluff his lines - he says “a breakthrough Internet communications device”, but the slide on the screen says ‘Breakthrough Internet communicator’. The audience reaction, particularly after the fervent response to the long-awaited announcement of an Apple phone, is muted.

Who could have foreseen the impact this third part of the trinity would have on our lives, our society, our culture? In 2007, there wasn’t even a word to describe a product that put the entirety of the internet in your pocket. Everyone knew what an iPod was, and mobile phones had been the most groundbreaking technology of the last decade.

But in 2007 nobody really used their phones to look at the web. That was something you did on a laptop or PC at home or at your desk. In an era of Nokias and Blackberrys, mobile phones were for texting or reading emails, not browsing websites.

The iPhone changed everything, not because it was an iPod or a phone - the two products that were merged to give it a name - but because of the third, that clumsy string of multi-syllabic buzzwords: ‘breakthrough internet communicator’.

The iPhone made the web ubiquitous. It opened up every single minute of our lives - all the places we went to, all the things we did - to the tentacled reach of information networks. But the internet wasn’t just coming to us, it was taking traces of us back along the same tentacled networks. As we carried it around, it created the opportunity to create data about us, tracking us not just in our homes and offices, but everywhere, all the time.

This wasn’t at all obvious when Jobs announced the iPhone from the cavernous stage of the Moscone Center. It didn’t feel like a surveillance device back then. That needed another innovation, one that was announced five months later, on a much smaller stage, about one and a half miles away.

The San Francisco Design Center is a collection of showrooms and a galleria, remodelled in the 1970s by design pioneer Henry Adams out of an old Ice House that supplied San Francisco’s fishermen. The main event venue is an atrium, flanked on two sides by four stories of showrooms, creating a tall, narrow box with a stage at one end.

With the right set and lighting design, it can be a stylish and even breathtaking space, but on May 24th it was almost completely dark, with a row of lights along the top of the stage illuminating a simple lectern with a microphone and an open laptop.

As Daft Punk’s Harder, Better, Faster, Stronger plays over the PA, a young man walked stiffly to the front of the stage to polite applause and a few whoops from the audience. Unlike Job’s signature black polo, he’s dressed like pretty much all the other coders and entrepreneurs who had flooded to Silicon Valley in the previous decade - faded jeans in an unflattering but comfortable wide cut, anonymous grey t-shirt, and a black North Face fleece jacket with the collar popped. His tousled mousy hair is in the process of outgrowing a Roman emperor style bowl cut, and his face is pale and flat.

This is Mark Zuckerberg, and he’s 23 years old. Facebook, the company he had started in his dorm room three years ago, has just reached 20 million users. He stood at the front of the stage, rigid with his chest puffed out, and looked around wide-eyed, wetting his lips multiple times before he speaks.

“Today, together, we’re going to start a movement.”

 There’s more polite applause and whoops, but nothing like the reaction Jobs had got a few months before. He takes a few steps around the stage, stiffly, head upright and eyes wide. His movement looks jilted and unnatural, like a character in a video game. He wets his lips again.

“At Facebook, we're pushing to make the world a more open place. And we do this by building things that help people use their real connections to share information more effectively.”

There are many ways in which Zuckerberg has articulated his vision for Facebook, but a few words tend to recur - open, connected, information - that in 2007 chimed with the optimistic and progressive vision for what was called ’Web 2.0’. If the first version of the web was about connecting pages through search, the second was about connecting us, through social networks.

In the early 2000s there were hundreds of sites online that promised to create a network of social connections for your online self, and then use those connections to make your life, somehow, better. Most of these sites centred on a particular demographic or media type - Bebo was for teens, Myspace for music, Flickr for photos. Facebook had built it’s early growth by focusing on college students, but Zuckerberg’s ambition was to do far more than that.

Facebook was about to reach an important tipping point. By the end of 2007 it aimed to reach 50 million active users, with only 25% of them college students. Facebook didn’t just want to connect students - it wanted to become the default social network for everyone on the internet. To do that, they need to build a platform that was not based around a specific demographic or media format. They needed to focus their platform on the core connections between all of us, and how information flows along those connections. They needed to focus on something called the social graph.

“We've always thought that Facebook is a tool that everyone in the world could use. Everyone has friends, acquaintances, connections, and through these connections, you really get a lot of information and really come to understand your world.”

Zuckerberg paced the stage as he talked, head stiff and upright, speaking quickly in short paragraphs separated by pauses, as if someone had told him to take his time, but he hadn’t quite nailed how to do this. There is absolutely no reaction from the crowd for most of his talk. He delivered the same kind of big lines that had caused wave after wave of applause at Steve Job’s keynote - ‘This is a big day for Facebook … Imagine all the things we’re going to be able to build together!” - but unlike the Apple crowd,  the response here is mostly silence. 

“People share information through their friends and the people around them in the world. So at Facebook we've constructed a graph of all of these real social connections that people have, and we use that to help people share information more effectively online, the same way they do in the real world using the same connections. The social graph is the network of connections through which people share information and communicate”

Zuckerberg then brings up a slide with a simple graphic - a white silhouette portrait in a Facebook-blue circle.

“We can start with a single blue guy. Now, this guy can go ahead and he can create a profile for himself, to represent his identity and have a place to store his media. But what's interesting here isn't a single person, but a relationship. Let's go ahead and add one of his friends. Now we have two people and they both know each other, and they can communicate offline, if one of them picks up the phone to call the other, or if they take some time to hang out or if they randomly bump into each other, but these methods are all synchronous, so they're kind of inefficient. In order for them to work, both people have to be paying attention to each other  at the same time. There's a better way to do this, and on Facebook, it's simple.”

On screen the blue guy has been joined by a yellow guy, connected with a blue pipe that shows lines of information flowing between them.

“On Facebook, these real connections become more efficient and people get more value out of all of their relationships. Now where Facebook really excels is in helping keep up with all their connections at the same time.”

The slide changes - the blue guy is now surrounded by a circle of yellow guys, each with a pipe connecting them to him, with more lines representing data flowing through the pipes.

“So if we continue extending this graph, you'll see that what we're building is a massive network of real connections between people through which information can flow more efficiently than it really ever has in the past. This is the social graph, and it's changing the way the world works. It's making the cost of communication between people so low that now information can flow between people through networks, faster than it can ever be pushed out by a few big companies. We're at a time in history when more information is available and people are more connected than at any other time in the past. And the social graph is at the center of that.”

In two presentations, just a few months apart, two products were launched that combined to create the ultimate engine for measuring the audience. One was a device filled with sensors and connectivity that exceeded even the most feverish dreams of the broadcast and audience research industry of the 20th century. Instead of having to sample data from a few thousand households, we were all creating data all the time, and not just in the living room, but everywhere we went.

The second was a powerful way of taking this data and creating a one-to-one scale map of how we are all connected. The media empires of the 20th century were built around sampling - using statistics to take tiny sets of data and extrapolate them across audiences of millions. Facebook didn’t need sampling to build a new media empire for the 21st century. They had data generated directly from millions -  and by the 2020s, billions - of people.

The problem of measuring the audience has been inverted - instead of having to use complex maths to extrapolate tiny data sets to represent huge audiences, we now use complex maths to reduce huge data sets until they represent tiny audiences.

Back in May 2007, as Zuckerberg is explaining this new world of social graphs in the San Francisco Design Center, there is still silence from the crowd. In a moment of irony, he’s giving a detailed breakdown of how Facebook will measure and monetise our every reaction, and he’s getting no reaction at all.

“So we have this powerful network of connections and it's growing really quickly. So what are we going to do with it? We're going to use it to spread information, spread information.”

Zuckerberg is speaking nervously now, and repeats this last phrase as if it were an affirmation: “spread information, spread information”.

“We've found that we can actually make information spread through the social graph. And here's how- if we can build an application so that for every person who uses it, they get more than one of their friends to use the same application, then that application will actually spread exponentially through the social graph.”

Still no applause. Zuckerberg is laying out his ambition to build a single network that connects the entire world, and he’s dying on stage. He needs to make it more real, to give an example that helps his audience understand his power. The illustration he chooses is, in retrospect, quite revealing.

“Here's an example. Recently, some people on Facebook have been using the groups application in order to organize support for political candidates. They go and they create a group - it's simple. And then they go and they invite some of their friends to join that group. Offline they might have had to create a club and bug some of their friends to join the club, but this is way more efficient.”

The slide now shows a network of blue couples - one vaguely male, one vaguely female - connected with the same pipes. In the top right, to represent the political group, there’s an American flag.

“So now all they need to do is select the people they want to invite and - done. Those people get an invite, and some of them can join. For those who joined it goes in their profile and maybe a newsfeed story gets published for their friends. Some of them see it, and they join, and it just continues spreading like that. And this is how we ended up with these groups on Facebook that have hundreds of thousands of members that are being used to organize political rallies. Both the group and the group's application spread exponentially through the social graph.”

The real power of Facebook’s new platform wasn’t just that it could connect the people on its network. In 2010, Facebook opened up their tools to let developers embed their like button and social graph all over the web. This new data from outside Facebook let developers do something remarkable - create a picture of someone even if they don’t give you any data. If someone wasn’t on Facebook at all, you could piece together a reasonable profile of them just by extrapolating data from people they were connected to outside of Facebook. Just by getting a few thousand people to give your app access to their Facebook data, developers could gather data on the millions of people in those users’ networks.

As Zuckerberg outlined the new Facebook Platform on stage, it looked like a culmination of all the technologies and ambitions of the audience measurement pioneers of the twentieth century. But with hindsight, we can see it was something even more powerful, even more personal. Arthur Nielsen Sr, the founder of the Nielsen company, once said to his son - ‘If you can put a number on it, then you know something.’ Zuckerberg was more ambitious - if you could get enough numbers, then you know not just something, but someone. Perhaps even everyone.

For nearly a century, audiences had been measured by capturing tiny amounts of data, and then, essentially, making up stories about what that data actually meant. Most of us were contained in one of these stories - sometimes called audience ratings, or audience research, or audience personas - but they weren’t really us, not on an individual level. They were imaginary archetypes, made up to help media companies make sense of our confusing and arbitrary behaviour. 

But now, like the picture on a high definition TV screen, we’d come into focus. After 2007, we were no longer personas or profiles. Our real lives  were there, spread out over multiple databases, diffuse but whole. We were no longer ghosts.

At the end of his talk, after demos from some of Facebook’s partner companies showing how they will use their new platform, Zuckerberg walked back on stage, applauding slowly to encourage the crowd, getting them warmed up for his big finish. He waited for the polite ovation to finish, then puffed out his chest again, lifting his chin up.

“This is the start of a really great day, and I'm really happy that all of you guys could be here with us to join us as we launch the next evolution of the Facebook platform. And to join us in making the world a more open place, one application at a time.

So there's only one last question: When are we going to launch Facebook platform?

And the answer is tonight.

So let's get started.”

In 2011, Stanford researchers Paul Thibodeau and Lera Boroditsky published research that showed how the way we talk about crime changes our ideas about what to do about it. They asked two groups of students to read reports about crime in their area - one using a metaphor of crime as a ‘beast’ that was rampaging through the neighbourhood, and one describing crime as a ‘virus’ that had to be stopped. Their research showed that students shown the ‘virus’ metaphor were more likely to favour policy that looked at the root causes of crime, such as social deprivation, whilst students who read the ‘beast’ metaphor story favoured enforcement policies.

In my day job, we spend a lot of time thinking about the metaphors we use to help shape people’s understanding of complex issues, and hopefully drive change. In fact, I know about the study above from a podcast we produced a few years ago called This Will Change Your Mind, looking at how the ideas that shape public thinking are developed and adopted. In my favourite episode, the hosts try and find out how the phrase ‘hole in the ozone layer’ emerged, a particularly successful climate metaphor that drove global efforts to cut CFC emissions.

It’s an odd metaphor, as there isn’t really an ozone layer, and the hole wasn’t really a hole, but the metaphor caught on with scientists, policy makers and the public. One of the reasons for its success might have been the visceral image of a hole in the earth’s atmosphere, and the associations of breached defences that created. This was the era of early video games like Space Invaders and Missile Command, where the player had to defend the earth by stopping alien attacks raining down. I was a kid in the 80s, and the hole in the ozone layer felt as terrifying as these pixellated invaders.

Back in the early 2000s, I was working at the BBC on a project to imagine what the organisation would do with our user’s personal data. At the time, there were only a few websites that asked you to create personal accounts, and most of the data we captured was relatively anonymous server logs. I developed a list of principles that I thought should inform the BBC’s approach to managing user data, and commissioned a service design agency to develop prototypes of what this might look like as products or services.

I was then asked to present this to the BBC’s executive committee, and gave probably the worst presentation of my life. It didn’t help that I started the presentation by explaining that this work might be important in a ‘post licence fee world’, before being softly chided by then Director General Mark Thompson that this wasn’t the place to discuss that kind of idea.

But more than that gaffe, I think the biggest problem was that I couldn’t really describe what personal data actually was. Not in a technical sense - the BBC ExCo in the mid-2000s wasn’t very technically savvy anyway - but as something that mattered, and was important to a vision of what public media could do in this new century. I had lots of fine statements about what we could do with data, and how it could bring value to our audiences, but the thing itself was immaterial, a poltergeist only visible through the things it moved.

Over the next decade, the most dominant metaphor for personal data ended up being ‘oil’. As the platform giants of Facebook, Google, Amazon and Apple built empires of products and services that tracked our every activity, data has been discussed as a vast, untapped resource, ideal for extraction and processing into cold, hard cash.

But in the last few years, there has been a backlash against this extractive metaphor. In 2018 Cory Doctorow described Facebook’s data as an empire of low quality ‘oily rags’, recasting the metaphor to one of industrial waste, not liquid gold. In a recent speech, EU Vice President Margrethe Vestager tried to reposition data as a ‘reusable resource’, a more ecological metaphor that suggests ways of extracting value that doesn’t pollute the public sphere.

All these metaphors imagine public data as a huge, passive, untapped resources - lakes of stuff that only has value when it is extracted and processed. But this framing completely removes the individual agency that created the stuff in the first place. Oil is formed by millions of years of compression and chemical transformation of algae and tiny marine animals (sorry, not dinosaurs). Data is created in real time, as we click and swipe around the internet. The metaphor might work in an economic sense, but it fails to describe what data is as a material. It’s not oil, it’s people.

At the moment the big platforms and governments are ramping up the battle over our personal data - who can collect it, what they can do with it, and where they can send it. But this is happening at a level far above our individual experience of data. The battle rages above us, like the missiles and aliens in the video games of my 80s youth.

The discussions around data policy still feel like they are framing data as oil - as a vast, passive resource that either needs to be exploited or protected. But this data isn’t dead fish from millions of years ago - it’s the thoughts, emotions and behaviours of over a third of the world’s population, the largest record of human thought and activity ever collected. It’s not oil, it’s history. It’s people. It’s us.

If you’ve been on the internet for a while - let’s say 5-10 years - you’ve probably felt the visceral kick of seeing someone or something in your data history that caused you pain. It could be Facebook’s ‘on this day’ feature sending you a memory of a traumatic event, or scrolling through your photo library to find a photo of a deceased relative or friend. Or it could be a moment of joy - the online store where you bough a much loved item of clothing, or that perfect gift for a friend.

After a year of lockdown, seeing reminders of life before the pandemic in our camera rolls and social media updates has felt especially melancholic. Groups of us cramming together in a bar or park to get into the shot, or hugging each other at a football game. That intimacy is what our personal data records, an intimacy that seems doubly ironic when it is played back to us, isolated in our homes, through the same devices we’ve relied on to connect us during the lockdown.

This is not a passive archive - these are records of how we live now, and how people live in our memories. They can be recalled with a touch, and brought back to life, even if they are bittersweet memories. We need metaphors for data that capture the agency and visceral emotions that our personal data can generate. Metaphors that link it directly into our lives and relationships, that help us recognise that this is us - we’re the ones being traded and sold and stored and analysed and processed.

Perhaps then we’d understand how we can handle this data in a more responsible way. A metaphor that puts our personal experience at the forefront will help us find out where to draw lines in how our lives are stored and processed, and to understand that the lines will need to be different for different people. I don’t know what the right metaphor is - memory and history are the concepts I’ve been mulling over, but they have already been used in computing in ways that blur and dull them.

Maybe we should be very explicit, and refer to data as our lives. Imagine if a service had to ask you permission to ‘track your life’ or ‘share information about your life with other providers’. Already that feels grittier, more visceral, than just ‘data’.

We urgently need to come up with metaphors like this, that bring the discussion over data down from the skies above us and locate it in the minutiae of our everyday lives. Because that, after all, is what this data actually is.

Part of the research I’m doing about measuring audiences is about the experience of being surveilled. I’m really interested in how companies got permission (or didn’t) from us to capture our lives as data, and what this trade feels like.

Writing this from 2020, we expect that pretty much everything we do is creating a data trail, but this shift in our expectations has happened very recently indeed. But even so, it already feels impossible to opt out - surveillance is the quid pro quo of modern networked technology.

Reading the history of audience measurement in the 20th century, I keep being surprised by how little data was generated, and how hard companies had to work to get these meagre morsels from us. The story I’m uncovering in the research is one of frustration from companies that we, the audience, continued to be ‘ghosts’ for far longer than they wanted us to. The battle to measure us, to turn all our needs, wants, choices and behaviours into data, was one that capitalism pretty much lost until the 2000s. Then the smartphone came along, and the dam broke.

I’ve started thinking about my own life, and how much I can remember, or reconstruct, the ways in which I have been measured and recorded. I feel surprised, strangely, that much of my life generated no permanent data. I’ve quickly assimilated a world in which my personal data is the key that starts pretty much any product or service. So it feels odd to remember that, for most of my life, I’ve been a ghost.

If you had an interesting history of personal data - if you were a Nielsen Family, or if your family filled out market research reports - I’d love to hear your story, so please let me know if you’re willing to have a short interview/chat for my research. Here’s what I can piece together about my own data life:

1972-1976
I was born, premature, on Feb 6th 1972. So there’s a birth certificate - the first record created that acknowledges I exist. A friend once described what it felt like registering his first child’s birth, mixing their lives into the state infrastructure in a way that is irreversible - ‘as intractable as milk in tea’. I love that phrase.

These are the records that are most likely to survive us. You can ask your GP for your medical records, and I wonder how far back they go, and what details they contain. I’m going to do this for my research, once it feels ok to bother them again with such a trivial task.

I am an identical twin, and many years later my mum mentioned she was asked (by whom? The hospital?) if she would be willing to join a longitudinal study about twins. She refused. If she had accepted, my brother and I would have been regularly recorded, analysed and published in research papers. But instead, we remained invisible. Chalk one up for the ghost.

1976-1990
I had an analogue childhood, although we were early adopters of home computers, and had a cup modem on our computer in 1988, dialling in to early bulletin boards and MUDs. No records of these early digital activities exist. The computers are long-gone, and the networks were a lot more forgetful then.

There are family photos, but these were recorded on long-lost negatives. Some prints are still around, but I’ve not published them anywhere. I haven’t scanned them for social media, unlike my daughters, who have already shared more photos with digital networks in their few teen years than I had taken, in total, in my childhood.

We bought things, but nobody knew it was us buying them. My parents had one credit card when I was a kid, but ran up bills on a foreign holiday and had to cut it in half when they got home. We bought newspapers (the Mirror for my parents, The Guardian for me when I was a teen, with pretensions to join the middle-class), but they were measured by returns at the shop, not in our living room.

We were measured in bulk, huge swathes of numbers recorded as vans were sent out with palettes of newsprint, returning with the unsold stock for the previous day. The delta between the two was us - the readers - counted by an absence as much as a presence. Our TV was always on, but we were never asked to contribute diaries or data for ratings. So all that attention - those hours spent watching Hill Street Blues, Cheers, The Young Ones - went unnoticed and uncounted.

We went to the shops to buy milk, crisps, meat, shoes, school uniforms, birthday and Christmas presents. The cash transactions were recorded at the tills, but again, it wasn’t us that was being recorded. Retail empires reached no further than the door of the shop - once we passed that threshold, we became ghosts again.

There were a couple of exceptions. My parents had Kays catalogues, and later the Next Directory, sent through the post. So there would have been records of goods bought on HP or the ‘never-never’, records of payments made and missed, letters sent with red labels on them. I assume none of this data still exists.

There was also the ‘Man from the Pru’, who used to come and drink tea on our sofa whilst my parents settled their life insurance. The Prudential started sending agents to our homes in 1848, and at the peak in the 1960s had over 10,000 agents visiting 6 million homes in the UK, each one with the trademark leather briefcase. Ours was very tall, and in my memories he looks like Richard Osman.

1990-1996

I left home in 1988 to go to Glasgow School of Art. There will be some records there, but the recent, heart-breaking, fires at the Mackintosh building will have destroyed some of them, including the hard-bound copy of my final dissertation, which is no great loss. My local council in Hertfordshire will have records of paying my student grant, and the newly-created Student Loan Company will track me borrowing the maximum amount every year, starting with £400 in 1990.

After Art School I spent a year on the dole, so there will be records of my visits to the unemployment office, and requests for housing benefit. The place I used to go to in Maryhill to sign on was one of the biggest in the country at the time. So my data will be a tiny speck in a universe of unproductive lives - people unable to work, unwilling to work, or, like me, trying to stretch the boundaries of what ‘work’ meant in the first place. Around the same time I did a couple of paid drug tests as a human guinea pig at Heriot Watt University’s medical research labs. I’m hoping the notes from these turn up in my medical records, and that there are no surprises.

1996-2003
The first few years of adulthood would have created more permanent traces - bank accounts, national insurance numbers, etc - but this gathers pace when I start work full time, first at two art centres in Yorkshire, then at the BBC. The moment I first used my BBC pass to enter the old Television Centre was a genuinely emotional moment. I had email for the first time, and there are still a couple of records of my early work email addresses in usenet and mailing list archives that are still online.

Up until now, pretty much every tiny trace of data about my life has been recorded for me, and mostly by the state. The early years of the web mark the time that I start to create a data trail by choice.

I’m there, in the archives of net.art mailing lists like nettime and Rhizome, announcing experimental digital art projects and taking my first stabs at writing about culture. Rhizome contacted me a few years ago asking permission to reproduce something I wrote in 1999 for their Net Art Anthology. I had no recollection of writing it at all, but there it was - clearly by me, in their archive.

And most of all, I started blogging. I had profiles on all the new social media networks at the time - ICQ, Myspace, Friendster, Bebo, etc - but these have all bit-rotted away, much faster than the paper records of the state. The first record at Test.org.uk is from February 2003, but it’s already referencing something I’d written elsewhere on the web.

The early 2000s were a period of rapidly filling boxes on the web with stuff - words and images of what we were thinking, feeling and doing. This is what we now eulogise as ‘the web we lost’, the optimistic era in which making ourselves visible to the network felt like a revolutionary, liberating act. The connections between us were made deliberately, with intent, as we subscribed to RSS feeds, added each other to blogrolls and created a network of trackbacks to each other’s writing.

2003-now
Meanwhile, the long-thwarted desire for capitalism to measure our every activity was about to have its moment. They’d got it wrong all along - instead of walking into our living rooms with a suit and briefcase, like the man from the Pru, they needed to create fake living rooms - places on the web where we would happily talk about our lives with anyone willing to listen.

Between 2003-2007 the tools to measure our digital lives grew quickly, first as analytics, then as algorithms, and then, with the launch of the iPhone, as hardware. After a century of being ghosts, we finally snapped into focus, and the irony was, we did it willingly.

It’s only thirteen years since the launch of the iPhone, and only a few days until Facebook, launched Feb 4th 2004, turns sixteen. I’ve been extremely online, and extremely visible to networks, for about the same amount of time I’ve been a father.

And just like parenthood, I’m finding it hard to remember what life was like before.

On May 9th, 1961, Newton Minow stepped up to give a speech to the National Association of Broadcasters at the Sheraton Hotel in Washington DC. Just a few months earlier, Minow had been appointed Chairman of the Federal Communications Commission, the organisation that regulated the US broadcasting industry, by President John F. Kennedy. Minow and JFK shared a vision for how television could impact society, worried about how their own children were spending so much time watching TV, and so when Kennedy was voted President, Minow asked if he could be appointed to the FCC.

The day before Minow gave his speech, Kennedy had spoken to the same audience. He was only a few months into his presidency, developing his vision of a US standing at the beginning of a ‘New Frontier’ - facing the challenges of communism and post-war political upheaval, but also the opportunities of new technologies and the space race. In his speech, President Kennedy told the audience of broadcast executives about his political vision for America, and how the new communications technologies they controlled would serve the fight for freedom against despotic rivals:

“…if we are once again to preserve our civilisation against our enemies, it will because of our freedom, and not in spite of it. This is why I am here to day with this most important group. For the flow of ideas - the capacity to make informed choices - the ability to criticise - all the assumptions upon which political democracy rests - depend largely on communication. And you are the guardians of the most powerful and effective means of communication ever designed. In the rest of the world this power can be used to describe the realities of communist despotism - and to give a true and responsible picture of our free society.

[…] And here, in our own country, your power is used to tell our people of the perils and challenges we face - of the effort and painful choices which the coming years will demand. For the history of this nation is a tribute to the ability of an informed citizenry to make the right choices in response to danger. And if you play your part - if the immense powers of broadcasting are used to illuminate the new and subtle problems which our nation faces - if your strength is used to reinforce the great strengths which freedom brings - then I am confident that our people and our nation will once again his to the occasion.”

Television was coming to the end of its first Golden Era, as American families adopted the new technology in record numbers, and advertisers flocked to the new networks. Regular network broadcasts only started 14 years earlier, in 1947, when there were 44,000 TV sets in US homes, around 30,000 of them in New York. By 1960, there were 52 million, one in almost nine out of ten households.

That rapid growth had led to problems. In the late 1950s, the TV industry was rocked by a series of quiz show scandals, as producers chased ratings by slipping answers to popular contestants. The FCC had introduced new regulations in response, and as the leaders of the US broadcast industry waited in the ballroom of the Sheraton hotel for Minow’s speech, they were anxious to hear what the new FCC Chairman was going to do next.

Minow went much further than even the most pessimistic broadcast exec could have expected. He started by reassuring them of his knowledge of the sector, reeling off a list of stats about how successful and profitable the networks were, but also recognising the anxiety they might have about his role:

“It wouldn't surprise me if some of you had expected me to come here today and say to you in effect, "Clean up your own house or the government will do it for you." Well, in a limited sense, you would be right because I've just said it.”

But Minow didn’t see his role as that of a teacher punishing a naughty classroom. He didn’t want to just issue a slap on the wrist, but to do something much more revolutionary.

For him, the problem with the emerging TV sector was not a lapse of morals about running quiz shows, but something much deeper - it lacked a vision for the public interest potential of this new medium. In his speech, Minow linked the need for a public interest vision to the ‘new frontier’ vision JFK had articulated the night before:

“Your industry possesses the most powerful voice in America. It has an inescapable duty to make that voice ring with intelligence and with leadership. In a few years, this exciting industry has grown from a novelty to an instrument of overwhelming impact on the American people. It should be making ready for the kind of leadership that newspapers and magazines assumed years ago, to make our people aware of their world.

Ours has been called the jet age, the atomic age, the space age. It is also, I submit, the television age. And just as history will decide whether the leaders of today's world employed the atom to destroy the world or rebuild it for mankind's benefit, so will history decide whether today's broadcasters employed their powerful voice to enrich the people or to debase them.”

For Minow, the landscape of television in 1961 was a long way from achieving this vision. He noted that in the networks’ schedules for the next season, amongst 73 hours of prime-time viewing, 59 hours were dedicated to sitcoms, quiz shows, action dramas or movies. This was the ‘vast wasteland’ that he wanted broadcasters to change:

“[…] when television is bad, nothing is worse. I invite each of you to sit down in front of your television set when your station goes on the air and stay there, for a day, without a book, without a magazine, without a newspaper, without a profit and loss sheet or a rating book to distract you. Keep your eyes glued to that set until the station signs off. I can assure you that what you will observe is a vast wasteland.

You will see a procession of game shows, formula comedies about totally unbelievable families, blood and thunder, mayhem, violence, sadism, murder, western bad men, western good men, private eyes, gangsters, more violence, and cartoons.
And endlessly, commercials -- many screaming, cajoling, and offending.

And most of all, boredom. True, you'll see a few things you will enjoy. But they will be very, very few. And if you think I exaggerate, I only ask you to try it.”

Minow asked the broadcast execs why so much of television was so bad. Was it pressure from the advertisers? The need for higher ratings and mass audiences? He didn’t believe the usual answer that ‘this is what the public wants’ - for Minow, the public interest was not just ‘what the public were interested in’. He had a keen knowledge of how TV ratings were captured, and challenged the broadcast execs to ask themselves whether ratings were a true measure of what the audience really wanted to watch:

“I do not accept the idea that the present over-all programming is aimed accurately at the public taste. The ratings tell us only that some people have their television sets turned on and of that number, so many are tuned to one channel and so many to another. They don't tell us what the public might watch if they were offered half-a-dozen additional choices. A rating, at best, is an indication of how many people saw what you gave them. Unfortunately, it does not reveal the depth of the penetration, or the intensity of reaction, and it never reveals what the acceptance would have been if what you gave them had been better -- if all the forces of art and creativity and daring and imagination had been unleashed. I believe in the people's good sense and good taste, and I am not convinced that the people's taste is as low as some of you assume.

[…]

You must provide a wider range of choices, more diversity, more alternatives. It is not enough to cater to the nation's whims; you must also serve the nation's needs. And I would add this: that if some of you persist in a relentless search for the highest rating and the lowest common denominator, you may very well lose your audience. Because, to paraphrase a great American who was recently my law partner, the people are wise, wiser than some of the broadcasters -- and politicians -- think.”

This was not just an airy vision - Minow then set out six principles that would guide his Chairmanship of the FCC, starting with the principle that the people own the airwaves, not the broadcasters, but also vowing to keep government censorship out of broadcast networks, and to stop fighting the still raw battles around the quiz show and payola scandals. And he went on to announce initiatives that he had already started to approve - an early experiment with pay TV, the opening up of UHF spectrum to allow more broadcast networks.

Unlike President Kennedy’s high political rhetoric the night before, Minow delivered not just a vision, but a clear plan of action. He was not asking for permission from the broadcasters, but challenging them to come on the journey with him.

Although his Chairmanship of the FCC only lasted a little over two years, his vision fundamentally changed the ‘vast wasteland’ of commercial broadcasting, and paved the way for public media in the US. He ensured that the new UHF spectrum would carry with it a requirement for educational programming, paving the way for the 1967 Public Broadcasting Act that created the Corporation for Public Broadcasting, and later PBS and NPR. Without Minow’s vision, there would be no Sesame Street, Frontline, Nova or This American Life. Public broadcasting in the US might seem tiny compared to the BBC and other public-funded broadcasters around the world, but without Minow, the US wouldn’t have anything at all.

We are almost exactly sixty years on from Minow’s speech, and we are at another inflection point for a new communications technology. We might not have the youthful optimism of JFK’s ‘new frontier’, but we are at a similar moment of rapid change, at the end of the first golden era of another global network. Television broadcast networks were only 14 years old in 1961 - social networks like Facebook and Twitter today are also still in their unruly teens. Anyone challenged to spend a day scrolling through their social feeds will find a ‘vast wasteland’ even more problematic than the one Minow saw on his television screen.

The last couple of years have seen a growing conversation about what we can do to change that vast wasteland, including breaking up the monopolistic tech companies, lobbying advertisers to stop funding bad actors, and many initiatives to invest in technology that starts with a public interest vision rather than commercial gain.

And finally, after years of growing pressure, the platforms themselves have started realising they are responsible for some of the content they host on their platforms. Banning the President of the United States of America from your platform is an unprecedented step, a crisis far more serious than the payola and quiz show scandals that TV faced in the late 1950s.

What we don’t have, and sorely need, is a vision from a leader like Minow of a better alternative, tied to a practical programme of change, and the political power to make it happen. The monopolistic technology platforms are deeply immersed in our everyday lives, as intractable as milk in tea. It feels impossible to imagine what a better version of them could look like. Many of us mourn the early optimism of the social web, but we can’t rewind time and try again.

What Minow did was not just identify the problem, but demonstrate a deep knowledge of the economic and technical issues facing the industry, and articulate a vision for how to make it better. He didn’t just criticise, but challenged his audience to stop hiding behind commercial arguments for the status quo, and start serving the public interest. Towards the end of his speech, Minow referred back to an earlier quote from Governor Collins, the President of the National Association of Broadcasters, about the need for broadcasters to serve the public interest despite the pressures of advertisers:

“I join Governor Collins in his views so well expressed to the advertisers who use the public air. I urge the networks to join him and undertake a very special mission on behalf of this industry. You can tell your advertisers, "This is the high quality we are going to serve -- take it or other people will. If you think you can find a better place to move automobiles, cigarettes, and soap, then go ahead and try." Tell your sponsors to be less concerned with costs per thousand and more concerned with understanding per millions. And remind your stockholders that an investment in broadcasting is buying a share in public responsibility. The networks can start this industry on the road to freedom from the dictatorship of numbers.”

We are living through the consequences of another ‘dictatorship of numbers’ right now. We thought the social web would lead to a ‘new frontier’ of progressive change and freedom, but instead we have monopolistic platforms run by algorithms that sell our attention and emotions to advertisers.

We need a bold vision for how we can change this - not a mournful paean to the failed dreams of 14 years ago, but something that ties a new vision of public value in the digital age to a practical plan for the projects, institutions, networks and regulations that can make it happen. This needs to come not just from activists and campaigners, but from someone, like Minow, with the power to bring it to life.

Like JFK in 1961, we’re about to see a new President in the United States, and a new set of appointments to the institutions that run public life. I hope at least one of them is a new Newton Minow.

Here’s a list of some of the live events I’ve seen this year:

Laura Marling at the Union Chapel
Lianne Le Havas at the Roundhouse
Nick Cave at Alexandra Palace
Billy Eilish live in LA
Pretty much every game the Boston Red Sox played this year
Every game Tottenham Hotspur have played this year
Barbershop Chronicles at the National Theatre
Twelfth Night at the National Theatre
A Midsummer’s Night Dream at the Bridge Theatre
A Streetcard Named Desire at the Young Vic
Hamilton at the Richard Rogers Theatre, NYC
Bruce Nauman at the Tate Modern
Andy Warhol at the Tate Modern

That would be a pretty good list of events any year, but of course, 2020 was an exception. I managed to see all these things whilst locked down, in my living room, because of COVID.

Some of these events I would have gone to anyway - I’d bought tickets to go to Bille Eilish in London with my daughter, and would probably have gone to at least one of the other gigs and a couple of Spurs games. But many of them I wouldn’t. I don’t go to nearly as much theatre as I’d like to, because of time, distance and money, and I only manage to see one or two Red Sox games a year if my US business travel coincides with them playing in the same city.

So as 2020 ends, and the vaccines start to give us a glimmer of hope that we could be near the end of the pandemic, I’ve been thinking whether this was a blip, or whether we’re entering a new era of remote culture.

A lot of my research on how we measure audiences has focused on a time almost exactly one hundred years ago, when radio networks were starting to grow their audiences. In 1925, US radio ownership was only 10%, but by 1935 is was 68%, and ten years later, 88%. This rate of adoption is not as fast as smartphones more recently, but much faster than colour TV adoption after the Second World War. Radio was the first time we saw mass adoption of a remote technology for streaming culture. Radio was the first time cultural entrepreneurs needed to think about the living room, instead of an opera house or theatre.

I’ve done a couple of talks recently about how COVID has affected audiences, and in this podcast for Substrakt I started using the phrase ‘remote culture’ to describe what’s happened this year. And that made we think there might be more parallels between the rise of radio in the 1920/30s and how COVID will affect culture in the next few decades. Both these moments posed new challenges to cultural providers - how do you deliver your stories into people’s living rooms? - as big a challenge for theatres and operas in the 1920s as it has been for COVID-hit venues in 2020.

The first services to live-stream into your living room predate even the radio. As early as 1876, just a few years after Alexander Graham-Bell invented the telephone, the New York Times was imagining how these new networks would revolutionise culture:

“The telephone - for that is the name of the new instrument - is intended to convey sounds from one place to another over the ordinary telegraph-wires, and it can be used to transmit either the uproar of a Wagnerian orchestra or the gentle cooing of a female lecturer […] When Madame Titiens is singing, or Mr Thomas’ orchestra is playing, or a champion orator is apostrophizing the American eagle, a telephone, placed in the building where such sounds are in the process of production, will convey them over the telegraph-wires to the remotest corners of the earth.”

By 1920, before radio networks started to take over, cities like London and Paris had successful subscription services like the Theatrophone or Electrophone charging around £5 for a year of culture delivered to your home - around £120 a year in today’s money, similar to the kind of prices charged by Netflix or Disney+. These ground-breaking services tapped into a previously unknown audience need for culture in your living room, something that radio and then television capitalised on in the second half of the twentieth century.

We could frame the last century as a battle between the living room and venue-based live events, as each new development of remote culture - the telephone, the radio, television, video recorders, the internet, smart phones - tempts us to stay at home instead of going out.

Radio was meant to kill live musical performances, and then television was going to kill theatre, or VHS kill cinema. None of this has happened of course - live experiences have visceral and social qualities that no technology has yet matched. But the uneasy balance between staying in and going out might have been tipped by COVID.

Early this month, Warner Bros announced that they will release their 2021 slate of movies - including blockbusters like Dune and Wonder Woman 1984 - simultaneously in cinemas and the streaming service HBO Max. This comes after 2020 saw Disney and other studios experiment with releasing films for home streaming first, and in the UK Cineworld shuttered all of their cinemas from October, some permanently. Just this week, Disney announced an incredibly ambitious slate of projects under their Pixar, Marvel and LucasFilms, with the emphasis far more on its home streaming service Disney+ than theatrical releases.

But the shift to remote culture is not just about blockbuster film and television. The cultural organisations that have done well under COVID were the ones who already had remote services up and running. The National Theatre in the UK had been streaming live theatre into cinemas very successfully for years, and after an initial free trial on Youtube, they’ve now launched a subscription National Theatre At Home service for £8 a month. London’s Soho Theatre’s home streaming service was established a few years ago, and they’ve now turned this into a production deal with SVOD service Amazon Prime. Organisations that haven’t done this, who have seen digital as just comms or marketing instead of a way of delivering high-quality services remotely, have struggled to make the same strategic shift.

Despite the hype that surrounds every new technology, audience behaviours actually change very slowly, but COVID has potentially introduce decades of change in one fell swoop. The lockdown has changed every sector - commercial property, transport, retail, entertainment, sport - and as we’re still very much in the middle of the pandemic, it’s hard to predict what behaviour changes will stick, and what ones we’ll gladly throw aside as soon as the vaccines allow us to.

Back in 2017, we published an online series on the next pandemic funded by the Gates Foundation, who have had pandemic preparedness as a strategic focus for years. In a recent podcast, Bill Gates set out seven ways in which the pandemic could change society, and one of the common themes is how COVID has quickly normalised ‘remoteness’ in so many areas of our lives. Many companies predict they will move to blended or fully remote workplaces after the pandemic, getting rid of their big city centre offices.

What will life look like when far fewer of us have to travel miles into big cities for work every day? What does a society look like that’s based around the home, not the office? How many things did we do because we were already in the city for work, and what things would we gladly commute to?

This all revolves around the most important strategic question we should be asking ourselves right now:

How does ‘remote culture’ affect what you do, and your relationship with your audience?

I want to strongly suggest that this is different from having a ‘digital’ strategy - digital technologies might well be a part of how you deliver your remote strategy, but we need to think more about the social and cultural contexts of the home. For many building-based organisations - whether this is a central office, a shopping mall, a theatre, museum or cinema - this means considering a future in which the building is no longer the sole focus of your strategy. Instead of seeing remote services as a side hustle to the main purpose - getting people through the door - the future might be an even balance between remote and building based provision.

There are huge potential advantages to embracing a remote strategy. Many of the cultural organisations we work with are saying audiences find remote services more accessible, either because they can fit them into the schedules of their daily lives, or because they remove barriers - distance, costs, or accessibility issues - that prevented them from using building-based services.

There’s also the potential of scale - audiences are now comfortable with paying for online content, so there is the opportunity to reach audiences that might never have come to a live performance or event. For their live-streamed gigs during COVID, Laura Marling and Lianne Le Havas grossed 3.5-4.5 times their average live gig income, from an audience that was 6-8x larger. Of course, this will pale in comparison to a full tour, but for mid-sized acts and theatre productions, remote performances will look like a viable option or addition to a traditional tour.

As I sit here, tapping away at the medium sized screen in my improvised office (the dining room table) before spending the rest of the day switching between the large screen in the living room and the small screen in my pocket, I’m really hoping that my 2021 list of events will involve some real live experiences. I can’t wait to see Spurs play at our amazing new stadium, or the RedSox at Fenway. I want to watch Hamilton again, live in the theatre, and hope that I can take my daughter to watch Billie Eilish live at the O2 in London.

But I’ve also loved the living room remote culture we’ve experienced this year. I’ve had some incredible cultural moments - and memories - with my family that we wouldn’t have been able to do in a normal year. It does feel like a tipping point, a moment in which our expectations and behaviours have been changed, some forever.

If we want to get audiences back into buildings after COVID, we’ll have to work harder at selling the social, visceral advantages that live events have over remote culture. Because the last hundred years has seen the slow, steady rise of remote culture. This has been driven by technology - the telephone, then radio, television, VHS and the internet - but it might be a virus that has finally tipped the balance in its favour.

The first phase of internet adoption was focused around the search engine, and the skill it taught us all was, broadly, how to be a librarian or researcher. Before the early 90s, most of us didn’t spend part of our day writing queries for databases and then evaluating and refining the results. But when we learnt to open a browser and type into a search box, that is exactly what we were doing. We’re all researchers now.

Something else has happened in the last 10 years, something that has given us another unexpected skill. After learning how to become researchers, we’ve all had to learn how to become schedulers. As smart phones and smart TVs have introduced streaming media and VOD services, we’ve had to learn how to manage our attention in ways that are incredibly sophisticated. The combination of these new skills, and the products we use to do it, are going to have a profound effect on culture and society in the next few decades, just as the rise of the browser and search engine did in the early 2000s.

I was thinking about this over the last few weeks as I analysed OFCOM data from their recent Media Nations 2020 report. Since the rise of internet adoption in the 2000s, live broadcast TV viewing has been dropping as a proportion of overall video viewing, but it’s always been above 50%. In the last six months, this has changed. Here’s the stats for 2019, when Live TV represented over 53% of video viewing:

And here’s the stats for April 2020, just after the COVID lockdown:

The interesting thing to note first of all is the huge rise in total video - 6 hrs 25 minutes in April 2019, more than 1.5 hours more than the average for 2020. April was unseasonably warm in the UK, but of course, we were on COVID lockdown, so were limited to 30 mins exercise outside a day. It looks like we spent that extra time watching video, but the big winner was SVOD services, which almost doubled from 34 minutes a day to 71 mins, whilst Live TV only added 24 mins, a 15.5% increase. This meant that for the first time, Live TV was less than 50% of our total video consumption.

But since then, it’s dropped even more. Weekly reach, measured as the number of UK individuals watching at least 15 minutes of broadcast TV, dropped to just below 80% of the UK population.

If we look at the figures for average minutes of Broadcast TV watched, you can see that 2020 briefly bucked the annual trend of decline as COVID hit, but is now dropping sharply to match the decline seen in 2019. Since 2014, average daily minutes viewing of broadcast TV has dropped year on year, losing around 45 mins a day over that period:

Why is this so important? Because we’re all watching way more video than ever before, but we’re watching it in different ways. We might be in the last decade of a cultural invention that has dominated media for the last century - the schedule.

I’m a bit obsessed with the schedule, as you’ll know if you’ve ever seen me talk at an event. It is a cultural construct that was invented at the turn of the 20th century as entrepreneurs working with new telephone networks imagined they might become a broadcast medium, and launched subscription content services piped to one-way phones installed in people’s homes. This gave them an interesting problem - how do you organise a programme of content when there are no physical limitations to how long you could keep something going? Their answer was to tie the lengths of their programming to the hours of the clock, and the broadcast schedule was born.

I think the broadcast schedule was one of the most culturally influential inventions of the 20th century. Although telephones didn’t end up as broadcast devices, radio soon came along, and then TV, both of them copying the schedule as a way of organising their content, and their audience’s attention. I’d argue that the power of the broadcast schedule (and the regulation of it by governments) created public spaces that deeply influenced the kind of politics, and progressive ideas, that affected society in the second half of the twentieth century. It’s only now, as the schedule is losing influence to algorithmic social streams like Facebook, that we can recognise exactly how influential the schedule has been on society.

When I worked at the BBC and Channel 4 in the late 2000s, I was surprised to find that the most powerful people in the organisation was not the talent or senior execs, but the schedulers. These people understood the dark arts of audience attention, and made decisions that could make or break a TV show, by running it against a competitor’s big hit, or pushing it backwards by an hour to a late-night slot. When audiences had little control over how to watch a show, the schedule, and the schedulers, were the hit-makers.

But now the power has shifted, and responsibility for managing our daily media attention has moved to algorithms and, well, us. We used to look at the Radio Times to see what was scheduled for our attention, but now we look at Screen Time to see how much attention we’ve given to our glowing glass rectangles. The really big trend of the last 10-15 years is just how much more of the day we now spend consuming media. This trend of how we spend our attention has been tipping towards digital for a while, but COVID seems to have been the watershed moment, and if lockdowns continue to curtail our behaviour for another six to twelve months at least, these new behaviours will only become more deeply embedded.

So right now, I’m fascinated with how we all manage our increased video watching habits, and how products have been developed to help, or hinder our efforts. I’ve written and talked a lot about algorithmic streams as a new cultural concept that will be as influential in the 21st Century as the schedule was in the 20th. Social streams have had a incredible effect on society and culture in the last decade, and they will continue to shape the next few decades. But something else has happened as well.

Alongside scrolling through social streams, we’ve also been binging through on-demand services. There is a whole series of design patterns around this behaviour that are nowhere near as well studied as Facebook and Twitter. This might be because they look, on the surface, quite familiar. Social streams were a radical departure for how we consumed content, atomising traditional structures like newspapers and magazines, and rearranging them in real time according to algorithms fed on data about our emotional responses. By comparison the electronic programme guides (EPGs) of VOD services like Netflix or Amazon are, at first look, a little like digital versions of the TV listings or Blockbuster video stores shelves we used to browse through in the 1980s and 1990s.

If the social stream has been optimised to give us an answer to the question ‘What’s happening now?’ (or more realistically - ‘What people are getting angry about now?’) then the EPG is being optimised to answer a more personal, less real-time question - “What shall I watch (or listen to) next?”

There are a couple of small, but important innovations in EPG/VOD design. I worked on the design of the first iteration of iPlayer at the BBC back in 2006, and the biggest problem we had was how to manage the egos of the different channel controllers on the iPlayer homepage. There was hardly any data collection in the product at that point, so algorithmic personalisation wasn’t really possible.

But there are other innovations outside of the algorithm. The first, started initially by Youtube before adoption by Netflix and all the VOD is auto-playing the next episode. This gets to the heart of the fundamental question we ask ourselves when we are self-scheduling our media - what should I watch next? This is solved for us in social streams, and YouTube, by an never ending scrolling list of new content, but social streams don’t seem to work in the living room in quite the same way. We seem to want to make more intentional decisions when we spend longer patterns of attention on video content, and social streams just don’t seem to work. Instead, auto-plays encourage us in one of the most important new behaviours we’ve learnt in the last few decades - binging. Binge viewing (or listening in the case of podcasts) has been one of the most important new behaviours in media, one that just would not have been possible in a world dominated by live scheduled TV.

The second interesting design feature is niche categories. Netflix has over 3,000 category codes in its EPG, each one representing a particular niche content area. These are mainly used as tags on content listings, but they’re a fascinating insight into how deep Netflix goes in cataloguing of niche content interests. For example, category code 1721544 represents Canadian Christmas, Children & Family Films; whilst 1133133 is for Turkish movies. This feels like a marriage of something like the Dewey Decimal System and the Folksonomies that emerged on early blogging and photo sharing sites like Flickr. Part of the skill of understanding what we might want to watch next is the ability to recognise patterns, and these deeply niche category codes feel like a really important part of how Netflix recognises patterns. They’re like Facebook advertising target groups, but not nearly as creepy. I don’t think they’re ever going to be a big part of the actual interface of an EPG for the user, but they’re a vital heuristic if you’re trying to train an algorithm. Netflix’s marketing is similarly aligned with specific niches and communities, with social channels like Strong Back Lead, Con Todo and Netflix Family.

The third interesting design feature is something that has caused a lot of controversy in social streams - trending charts. On Twitter and Facebook, trending topics are a kind of snapshot of society’s id - a crazed and manic combination of breaking news, celebrities and memes that feels like it’s being shouted into the sky, accompanied by a crowd with pitchforks at the ready. Netflix has only recently added a trending shows interface, and intriguingly, they show it as a top 10 chart. The chart, as a way of organising content according to it’s current popularity, was invented by the music industry in the 1950s, but has waned in influence in recent years as music listening moved to streaming services. Reimagined by Netflix, it acts as a useful window into what everyone else is watching at the moment. It doesn’t feel as real-time as trending topics, but instead gives a feeling of currency and shared experience that used to be the experience of watching scheduled TV. The trending chart is a little window into what everyone else is watching - a small watercooler moment into the wider Netflix audience’s behaviours.

None of these design elements is outrageously complex or innovative. In fact, compared to the relentless product innovation of social stream platforms, VOD EPGs feel like they are relatively stately in their development. But this is why they’re fascinating to me. They’re reinventing another part of our public realm - as the stream reinvented newspapers and magazines, VOD EPGs are reinventing TV channels and schedules. We need to be interrogating the design choices the streaming platforms are making, and asking ourselves whether they are creating opportunities for us to binge on diverse new story worlds that open up the world to us, or whether they are focusing us, like social streams, into ever more limited filter bubbles.

The last episode of the last season of this newsletter (and yes, I’m going with seasons as a publishing rhythm) came out in late February this year. Which really feels like not just a different time, but a different world. After 4 months under a never-breaking wave of unimaginable change, I feel like new rhythms of daily life are starting to emerge. They are fragile and precarious, but they’ll do for now. I hope you are finding your own new rhythms.

Over the last four months I’ve done a few podcast and zoom webinars for different people, including a talk about formats for Digital Works, masterclasses on attention and formats for Columbia University and Channel 4/National Film and TV School, and a good chat with my fellow Brightonian Richard Gillis for his Unofficial Partner Podcast about attention, Arthur Nielsen, and fandoms in sport.

I picked up a lot of new subscribers from that podcast, so welcome sports fans! Thanks for subscribing. This newsletter is very irregular, but very similar to the kinds of chat we had in Richard’s podcast. You’re joining 331 other subscribers to this list. I hope you stick around and enjoy it, and I love getting feedback and questions, so do let me know what you think.

In February I was gathering research material to try and understand how audiences reacted to the growing surveillance of their lives as media measurement tools developed in the 20th century. One of the areas I’ve been researching is the introduction of the ‘people meter’, a new way of measuring audience attention that emerged in the late 1980s.

As multichannel TV and VCRs took off, advertisers and broadcasters realised that their traditional models of audience measurement wasn’t capturing new audience behaviours. At the time, there were two complementary models for audience measurement - set top boxes and diaries. Set top boxes measured activity from the TV itself, sending data back down the phone line whenever it was switched on, or a channel changed. Diaries captured demographic behaviour, asking viewers to record their personal viewing every day.

Both methods had problems. Set top boxes didn’t require users to do anything different in the way they watched TV, so the data captured was complete and delivered instantly over the phone lines. But it only measured the TV, not the people actually watching it. Diaries gave an insight into personal behaviour, but they required users to physically fill in a diary recording their viewing, and it took weeks for the ratings companies to gather and analyse the data.

The combination of the two was imperfect, but at least they provided both detailed numbers and demographics, from which you could develop a model of what the audience might actually be doing. At least, you could when TV viewing behaviour was relatively simple. Cable, Satellite and VCRs changed all that.

I’ve been reading two books by Ien Ang, a Professor of Cultural Studies at UWS in Australia, that capture this moment in time really well. They were written in the 1990s, and so have fantastic access to contemporary magazine and journal articles about the rise of the people meters. She includes this great quote from an 1985 article by David Poltrack, vice-president of research for CBS, talking about how audience measurement was getting harder:

“It used to be easy. You watched M*A*S*H on Monday night, and you’d put that in the diary. Now, if you have thirty channels on cable you watch one channel, switch to a movie, watch a little MTV, then another program, and the next morning with all that switching all over the place you can’t remember what you watched.”

The people meter was introduced to solve this problem by AGB in 1987. The people meter was a development of the set top box device, but with the addition of numbered buttons representing each person in the household, plus a couple of extra ones assigned for guests. When someone started to watch TV, they pressed their button, and when they stopped or left the room, they pressed it again.

This was pitched by AGB and Nielsen as a perfect combination of both diaries and set top boxes, but there was a lot of resistance. One of the biggest problems for broadcasters was that the ratings it produced were lower than the previous methods. Early tests of the people meters showed ratings that were constantly 5-10% lower than the old diary and set top box system. When Nielsen ran the new and old systems side by side in August 1987, the people meter showed 1.3m fewer people watching TV.

Broadcasters were reluctant to accept the results of the new system, and critics argued that it was too similar to existing diary systems, which relied on people remembering to participate. NBC’s Vice President of research William Rubens dubbed the people meter ‘an electronic diary’ requiring people to ‘push buttons instead of pencils’, saying:

“People meters go against human nature. You can’t expect people to work on data entry during their leisure activity of watching TV. Either they take a leisurely approach to data entry, or TV viewing becomes work - and they may ease the burden by watching less”

This touches on an area of research that I’m particularly interested in - the tension between advertisers need to turn our lives into data, and the limits of our willingness to do so. In the last newsletter I mentioned the late 19th century worker’s slogan “Eight hours for work, eight hours for rest, eight hours for what we will!”. I’m really interested in how, over the course of 20th century, the ‘eight hours for what we will’ went from being invisible leisure time to valuable commercial data.

People meters eventually became the standard measure of TV audience in the US and most of Europe, although Nielsen continued using paper diaries until 2018. But the problem of how to measure audiences without relying on them do anything remained. The ultimate goal was a system of ‘passive measurement’ - an automated way of gathering data about our attention that didn’t require us to break the spell and turn that attention into work.

In Ang’s book Desperately Seeking The Audience, she quotes an article from TV Guide in 1984 discussing the ways in which Nielsen and the ratings industry was trying to solve the problem:

“One suggestion is to implant tiny electronic ‘bugs’ in the navels of all family members in a people-metered household. That way, the meter will automatically ‘know’ ‘who’s watching, with no action require of the viewer. Another solution is to give every family member a special bracelet or wristwatch that would transmit a signal identifying the wearer to the meter. Or how about an ultrasonic device (like those used or burglar alarms) in all the rooms with TV users, so that family members would be recognised by the meter the instant they switched on the set. Another modest proposal is to build into the TV set a photoelectric eye that would watch you. And finally - the device audience-measurement theorists only all ‘the whoopee sofa’: a divan wired to detect tiny variations in the temperatures of household members’ bottoms and thus identify them for the meter.”

It’s interesting that article was written in 1984, of all years. George Orwell did, after all, predict that TV would watch us many years earlier. But what I find fascinating is that all these methods suggested by Nielsen and the other ratings companies now read like prototypes for the connected home or internet of things devices we we’re being sold in the 21st century. The ‘special bracelet or wristwatch’ is clearly the Apple Watch. Although the Ring camera uses video rather than ultrasonic technology, the mention of burglar alarms makes the link too close to ignore. And earlier this year, Stanford University released research about a smart toilet’ prototype that recognised you by an ‘analprint scan’. We haven’t seen bugs inserted into our navels yet, but I just happened to watch The Matrix last night with my family, and a bug like this plays a significant (and quite gross) role early in the film.

Nielsen did test facial recognition technology in 1989 to replace the button-pushing in their people meter, but it just wasn’t good enough to work at the time. This is why I’m enjoying researching this period of audience measurement - the desire to track people every waking minute was there, but it needed the internet, and in particular the smart phone, to make it possible.

What we now call surveillance capitalism didn’t start with the rise of Google and Facebook - it was already there as an idea in the 20th century, but just nowhere near as widespread and sophisticated. Nielsen measured a couple of thousand households to represent the US public. In 2020, Facebook has over 235m US users, each one of them giving the kind of minute by minute data about their attention that was impossible in the 1980s.

Two kinds of innovation were necessary to make surveillance capitalism possible. One was technology, and the rise of smart phones and the web solved this. The second was more unexpected - a social revolution that made giving data about yourself feel like fun instead of work.

In 1980, Nielsen had to send people actual dollar bills in the post to get them to fill in their diaries. In 2020, we do it ourselves every day for free, and the money gets sent to Facebook instead. We are all people meters now.

There’s been a little flurry of subscriptions after I broke my hibernation last week. There are now 243 of you, so thank you all for your interest and feedback. If you’ve been sent this email and you’re not a subscriber, you can join by clicking on this big red button below. Thanks!

Subscribe now

Last week I mentioned that I was interested in stories about resisting surveillance in the home, and Michael Newberry sent me this excellent NYT article about an anti-Alexa bracelet. It emits ultrasonic noise, which will not only stop Alexa from listening to you, it’ll also annoy the hell out of any dogs and teenagers in the vicinity. The fact it’s a bracelet is nice coincidence, as there’s a long history of dubious medical bracelets, like the Power Balance Band, which had a strong following amongst elite athletes in 2011 before the company went bankrupt in the face of multiple lawsuits.

I’d like to imagine someone wearing the anti-Alexa bracelet on one arm, and an Apple Watch on the other. Surveillance on one hand, Privacy on the other, like Radio Raheem’s Love/Hate knuckle rings in Spike Lee’s Do The Right Thing - “The story of life is this - one hand is always fighting the other hand”

I’ve found a rich vein of research in the last week, and have quite a few old books about the history of TV ratings arriving in the post. One of the books is an academic essay collection called Television Audiences Across The World: Deconstructing the Ratings Machine, edited by Jérôme Bourdon and Cécile Méadel. The bibliography alone has led me to many great research sources, but the intro is also a really good short account of the politics and sociology of TV ratings.

Bourdon and Méadel suggest a number of ways of thinking about what we’re actually doing when we measure TV ratings. First, they refer to a 1977 essay where Dallas Smythe uses a Marxist framework to suggest that audience’s viewing activity should not be considered as leisure time, but as ‘work’ for capitalism to extract. This reminds me of the slogan from the late 19th century workers’ rights movements - “Eight hours for work, eight hours for rest, eight hours for what we will!”

I like the almost petulant vagueness of “eight hours for what we will!” The phrase not only stakes out the right for leisure time, it refuses to give any details of what that leisure might involve. It’s almost as if these workers knew that the 20th Century would increasingly surveil them, even in their leisure time, and they were pre-emptively turning themselves into ghosts. By the time Dallas Smythe writes about communications being the ‘blind spot’ of Western Marxism in the 1970s, the game is up, and capitalism already knows too much about what we do in our leisu

This leads to the second way of thinking about audiences - as a commodity produced by the act of measurement itself. In this reading, audiences are not a specific group of people, but are rendered as a set of data, used to facilitate the sale of advertising between broadcasters and brands. This reading puts the emphasis not on the activities of the audience, but on the politics of the measurements themselves - the metrics, not the masses.

Most Western countries use an independent agency to do this, like BARB in the UK or Nielsen in the US. The independence and continuity of how this data is produced is important to everyone involved in buying and selling advertising. If the methodology is questioned, it threatens the trust everyone must have to make the market function. I’m researching a couple of interesting moments when this trust was questioned in the US, such as the 1963 Harris Committee on broadcast ratings, and the 2004 PeopleMeter controversy. This last one is a particularly juicy story, involving Rupert Murdoch’s News Corp; the Nielsen Company; Hilary Clinton; a lobbying group who also worked for Al Gore and Michael Bloomberg; and a pseudo grass-roots organisation called ‘Don’t Count Us Out!’ who argued that TV ratings undercounted minority audiences.

This in turn leads to a third reading - measuring TV audiences gives the public a ‘voice’ that can be used by broadcasters and governments as a proxy for public opinion. As I explored in an earlier newsletter, the invention of broadcast media ratings in the early 20th century happened alongside other new techniques for measuring the public, like George Gallup’s invention of opinion polling.

For public service broadcasters in particular, ratings were not about rendering audiences as data for advertisers, but to create a ‘public’ that they could consider in programming and scheduling discussions. This feels familiar from when I worked at the BBC, where the audience research team regularly used ratings research to create composite personas of the BBC audience for BBC execs. Later on, when I worked at Channel 4, the commissioning team would meet every Thursday lunchtime to go through the last week’s TV schedules, and ratings were often invoked to discuss what ‘the audience’ had liked or ignored that week.

The issue with this third reading is that ratings are samples - they do not measure everyone equally. There are voices and opinions that aren’t represented by the ratings - in fact, one chapter in Bourdon and Méadel’s book looks at TV ratings in India, which only measures the ‘A’ group of India’s caste system, which is less than 10% of the total population. If ratings are used to indicate what ‘the public’ like or don’t like, then who does this ‘public’ include, and who is left out?

The story of TV ratings in the 20th century is the story of these multiple and competing readings of what an audience actually is, and who gets to decide which is most important. Does measuring our attention turn our leisure into yet another form of ‘work’ for capitalism to commodify? Or is it an abstract representation that needs to be agreed and trusted so that the advertising market can function? Or does it create a representation of the ‘public’ that can be used as a proxy for decision making by broadcasters and governments?

The tension between these three readings was one of the reasons why the job of measuring ratings was given to independent bodies. Ratings were too important to leave to either the broadcasters, the advertising industry, or the government alone. They need to be kept at one remove, so everyone could share the same illusion that they were true, honest and impartial.

In the 21st century, this tension finally started break, as Google and Facebook created advertising markets in which they were simultaneously the creators, the measurers, and the sellers of audiences. The digital platforms of the last few decades have profoundly broken the illusion that measuring audiences can be an impartial and arms-length act.

In doing so, they’ve broken Bourdon and Méadel’s three readings of the idea of an ‘audience’ and created a new hybrid definition. Our leisure time is no longer invisible to capitalism - instead, it’s the core resource for new forms of platform (or surveillance) capitalism. There is no longer a single metric for our attention, it is now a commodity that can be bought and sold in endless fractal formations to highly-targeted advertisers. And they’ve shattered the idea of a ‘public’, using our attention data to instead present the world back to us in individual, personalised streams curated by their algorithms.

We are no longer ghosts, but I don’t think we’re an audience or a ‘public’ either. We need to find a new language to describe what we are when we use digital platforms, or risk other people doing it for us.

There was a great interview with Malcolm Gladwell in the New Statesman last month, which slightly rehabilitates him after the brickbats he got for his last book. I love this quote most of all, about the challenge of writing popular books about ideas and history:

“At some point I’d like to write a book that didn’t have any theory. If I could write a book like Michael Lewis’s, where the ideas are there but exist entirely within the context of character, and the intellectual part recedes entirely into the background – that is the gold standard for me.”

This is a really high bar to aim for - the kind of book that is brimming with new ideas, but is driven purely by protagonists and action, not exposition and theory. The kind of book that gets made into a movie starring Brad Pitt, rather than the kind of book people buy because they feel like they should read it, but never actually get around to it.

Back in December, I took my levels of book writing procrastination to a new level. Not only was I not writing my book (or this newsletter), but I accepted an offer from a friend for some paid work doing a development edit on his book. It’s a book about some of the same themes - the media industries, disruption, and attention - but comes from far more of an economics perspective. But as writing avoidance tactics go, accepting his offer was a new level. There should be a saying - if you want to get some writing done, ask a writer who is struggling with their own book.

On the positive side, editing his book has helped me rediscover the discipline of writing, and how much more I enjoy editing to writing. Getting the stuff on the page for the first time is by far the hardest part. If I could pay someone to take all of the half-formed thoughts in my head and put them into Google Docs, I’d be very happy spending the next few months polishing the words till they shine. As we’ve discussed the book over lunched in Brighton, I’ve really enjoyed starting to recognise the structures, metaphors and callbacks, helping him tease these out to make them more like a Gladwell, and less like an academic paper. There’s a fantastic book sitting there, and it’s been really satisfying helping to pull it into shape.

Finishing the development edit this week has given me the impetus to dive back into the research for my book. At the moment, there are five questions that I’m interested in, and I’m starting to do the deep research to answer them. Over the next few months, I’ll share what I’m finding in more of a scrapbook style on this newsletter. I started this newsletter as if I was writing fully formed articles, but as I’ve subscribed to more Substack newsletters, I’ve preferred more informal and open-ended writing. Newsletters that feel like early blogging - an open and lightly formed exchange of ideas - are more fun than an Atlantic article in an email.

I’d like to turn this into more of a conversation. So here’s five things I’m thinking about at the moment. If you’ve got thoughts on this, or leads for potential books or sources to research, I’d love to hear from you.

1 - When did we first let companies listen to us in our living rooms?
I’ve been researching the history of surveillance, and am looking forward to Brian Hochman’s forthcoming book, but a lot of research is about state or criminal surveillance. I’m interested in how we started letting companies listen to us, as part of their need to track our behaviour and sell it to advertisers.

2 - How did companies imagine the living room in the 20th century?
In an earlier newsletter I wrote about James Robinson’s excellent research on how newsrooms and journalists imagined their audience. Having worked in large media organisations, I know there is a kind of folk memory of the audience at the heart of mass media. When the audience is invisible, we have to imagine them, and I’m curious how different groups - creatives, advertisers, regulators, etc - thought about the audience and what we were doing as they were trying to get our attention.

3 - How did regulation stop companies surveilling us in our homes?
Most of the writing about surveillance is about public space. There’s a lot of academic research on surveillance in cities and CCTV, but less on surveillance in the home. I’ve found a few interesting recent articles about Ring cameras, and a few piece on the rise of the Nanny Cam at the end of the last century. I’m curious about whether commercial surveillance in our homes was regulated as much as surveillance of us in public. I don’t think it was.

4 - How have we imagined and responded to surveillance in our homes?
Just as companies have to create stories and myths about what we do in our living rooms, so we - the audience at home - have created myths about what technology knows about us. There is a vivid history of psychological conditions mapping on to the technologies of the time. First, we thought the telephone was listening to us. Then it was the radio and the television. Now it’s Alexa and Siri.

5 - Why do we talk about public media when we consume it in private?
This is something that occurred to me as I’ve been reading Paul Starr’s magisterial ‘The Creation Of The Media’. This is an epic account of how media technologies have created public space over the last few centuries. But after telegraph, which was never a widespread domestic technology, most of the others he discusses - the telephone, radio, television and the internet - were mostly consumed in domestic settings. In other words, we talk about public media as if it was an agora - a public space that we travelled to for debate and learning. But really, the history of public media is the opposite - for the first time in history, the agora came to us, in our living rooms.

Alongside all this is my continuing effort to research the life of Arthur Nielsen, and his role in all the questions above. If there’s a chance that my book could be more like Gladwell’s description of Michael Lewis’ book - driven by narrative rather than theory - then Nielsen is the most likely protagonist at the moment. The friend who asked me to edit his book has put me in touch with someone who might be able to let me have access to the Nielsen archives. Fingers crossed.

In September 1964, LIFE magazine published an article about a new technology that tried to solve the problem facing all broadcasters - who is actually watching or listening to their shows? The technology was the Tanner Electronic Survey Tabulator, or T.E.S.T truck. Invented by electronic entrepreneur Jim Tanner, it was a van fitted with seven television sets, each tuned to a different TV channel. As the van drove through a neighbourhood, a circular antenna picked up the frequencies emitted by televisions in people’s living rooms. If they matched the frequencies from the TVs in the van, an indicator pinged and the correct channel was noted in the Tanner Survey for that house.

Tanner proposed it as a better alternative to the standard Nielsen ratings. At the time, Nielsen measured around 1,100 carefully chosen households, but Tanner claimed that with 300 T.E.S.T trucks he could measure 20 million TV sets in any half hour of the TV schedule. TV execs argued that they didn’t just need to know that TV sets were on, but who was watching - which family members, for how long, and the age and demographic info about the families. Tanner’s T.E.S.T truck couldn’t do that, so he offered a follow up survey so he could match the frequency recordings with demographic data.

Despite this, the T.E.S.T didn’t take off, although in the UK similar technologies were used for TV Detector Vans, checking to see that households watching TV owned the correct TV license. There is some doubt about whether these license detector vans actually did patrol the streets of the UK, or whether they were just a marketing device to strike fear into people without a TV license. But TV Detector vans did exist in the UK, dating back to 1952, whether they were regularly used as enforcement or not.

There are early versions of UK TV detector vans, like this example from the late 1950s/early 1960s, that look like they have similar technology to the T.E.S.T’s circular antenna:

Like a lot of technology in the 50s and 60s, these vans feel simultaneously futuristic, yet rooted in the everyday. The huge antennas and banks of TV screens speak of cutting edge technology networks, but the vans themselves are ordinary, usually seen doing more mundane tasks like taking school-kids to and from football matches. They are at once homely and strange - the very definition of uncanny. In his essay on the uncanny, Sigmund Freud uses the German word Unheimlich, which translates as ‘un-homely’. The TV Detector vans were definitely ‘un-homely’, creepy symbols of a paranoia that people had, and continue to have, about domestic technology - that we don’t just watch it, but it watches us back.

The story of how we’ve measured media audiences is, to a surprisingly large degree, a battle to find ways into our living rooms. Audiences became ghosts not because we disappeared, but because technology emerged that took media and culture into our private spaces.

Once we stopped going into a newsagent, bookstore, cinema or theatre to give someone money in return for culture, we became invisible. We stayed at home, and in order to continue making money, companies had to convince us to let them in (I’ll leave the obvious vampire reference to the reader).

The first device we let in our living rooms to measure us was the Audimeter, invented at MIT, but bought by Arthur C Nielsen to help grow his emerging radio measurement service. You’ll know if you’ve read my other newsletters that I’m fascinated by Nielsen, and think he’s the most influential, but overlooked, figure in the 20th Century.

The Audimeter was a metal box, with a slowing rolling coil of tape inside, onto which were etched marks triggered by a listener turning and tuning their home radio. I wonder what it must have been like to be one of the early radio audiences asked to take in an Audimeter. First of all, this magic box called a radio is invented that brings voices and sounds from all over the world into your living room. Then someone asks you to take ownership of this small metal box, about the size of a hardback book, and explains that it’s going to record everything you do with your radio, to be sent off somewhere for someone to analyse.

I wonder why you would do this. At this point in history, the living room was a very private space. Not every house had a phone or a radio, so domestic space was truly unknown and unconnected to the outside world. The only communication was printed - books, magazines and catalogues coming in, letters coming out.

Electricity brought networked communication into our private domestic spaces - first the telephone and radio, then televisions, then computers, and finally the internet. With every new device, there have been legitimate concerns and paranoid delusions about whether we are being listened to, and yet this doesn’t stop us inviting them in.

The Audimeter - the first device that actually did listen to what we did in our living rooms - was invented in 1936. Seventy-Five years later, Amazon announced the Alexa - a device that combined the broadcast capabilities of the radio and the data recording capabilities of the audimeter into one neat package, about half the size of a hardback book.

The smart speaker has been the breakout technology of the last few years, but it really is just combining two things that have been going on for seventy-five years - broadcasting audio, and listening to our living rooms. The difference now is that the listening and broadcasting are combined and sold to us in a single device.

Somehow, we seem to have come to terms with people listening to us in our living rooms. We’ve given up the privacy our our domestic space, and invited people in to have a listen. The history of how we’ve gradually invited in these spies is not well known, and it’s something I’m researching at the moment. I’m interested in how this happened, what worked and didn’t work, and when, and how, some people resisted.

In the end, it turned out that the best way of listening to what we do in our living rooms wasn’t to drive outside in creepily modified vans bristling with antennae. It was to package the microphone and speaker into one slick black cylinder, connect it to the internet, and ask us to pay for the privilege.

As with the hashtag and @-replies, Twitter noticed this emerging behaviour, and coded it into their product.

Let’s stop for a minute and focus on this process, as it’s an example of one of the most critical new developments in media networks in the last few decades:

Twitter builds a service that lets audience freely communicate with each other. It has limited functionality, so users start inventing shortcuts and hacks to add features to the service. Twitter starts to notice and track this behaviour, then decides to include it as a product feature.

This is a good thing, right? There are many design books on how to focus on emerging user behaviours to get insights into their needs, and how this can make your product better. So, changing the process from manually copy and pasting a tweet into a new tweet and adding ‘ReTweet’ or ‘RT’ into a simple tap of a button is a good example of user-centred design practise. Isn’t it?

Perhaps not. In the Buzzfeed interview, Weatherall describes his regret at how the retweet button affected Twitter as a public space. He recalls how the Gamergate controversy of 2014 - in which female games journalists were subjected to horrifically offensive and sustained criticism and trolling - was the moment he started realising the negative effects of the retweet:

“It was very easy for [the Gamergaters] to brigade reputational harm on someone they didn't like. Ask any of the people who were targets at that time, retweeting helped them get a false picture of a person out there faster than they could respond. We didn't build a defense for that. We only built an offensive conduit […] It dawned on me that this was not some small subset of people acting aberrantly. This might be how people behave. And that scared me to death.”

I’m interested in Weatherall’s insight that the retweet button was an ‘offensive conduit’, for which Twitter had no defense. When the retweet button (and soon afterwards, Facebook’s ‘share’ button) were added, most commentary was hugely positive. These platforms were recognising the social behaviours of their audience and turning them into code, massively increasing their ability to amplify messages that are gaining popularity, regardless of the audience reach of the original creator.

No-one imagined at the time how this could have negative consequences. As Weatherall recalls, “Only two or three times did someone ask a broader and more interesting social question, which was, ‘What is getting shared?’ That almost never came up.” Twitter focused on users’ behaviours, but were blind, or at least uncurious, to the context and consequences of the behaviour they were automating.

Before the retweet and share buttons, it was hard for someone with a small network to get their content out into very large groups. Ideas and memes went viral, but the process involved a lot more work - they needed to be shared manually within communities, often getting remixed and developed as part of the work needed to break them out into larger networks.

Kenyatta Cheese, founder of NYC social agency EA1, has given a number of excellent talks about how communities remix and adapt memes as they spread. In fact, he was one of the founders of Know Your Meme, a site dedicated to tracking the rise, adoption and adaptation of memes across the internet.

In his brilliant talk ‘The Story of the Gif’, he describes how the technically limited Gif format ended up outlasting the more sophisticated Flash format, precisely because it made the work needed to create moving image memes more accessible to a wider audience. Gifs ability to created short loops of video into a highly compressed format, together with its compatibility with basic HTML, made it hugely popular as a format for turning short clips of popular culture into a form of communication. Ground Zero for this behaviour was Tumblr, which had two critical features that encouraged Gif usage:

“The first is that you can upload a group of photos into a single “photoset” which means that you can get really creative on the timing and juxtaposition of animated gifs.

The second is that the heart of Tumblr is the “reblog” which is more than just a retweet or a share — it allows you to add your own text and images in response and when you reblog it, your response is just as important as the original post. This facilitates discussion. And because reblogs show up on the original post as “notes” it means that you can actually see what other people have written in their responses.

So if someone were to watch the Disney movie Snow White, she may love it so much that she wants to share it with their followers on Tumblr so she creates a gifset. She gives it tags so that it’s findable by other people exploring topical content on tumblr. The post gets reblogged, perhaps with commentary on how much they also loved the movie and what it meant to them.”

Cheese’s insight into how Tumblr’s ‘reblog’ button differed from Twitter’s retweet and Facebook’s share buttons is really important. Tumblr encouraged user comments on reblogged posts, and tracked them on the original post. They made the conversation as important as the original content.

When Twitter hard-coded retweets into the platform, they made it harder for users to add comments to a retweet - it was easier to just share the tweet without context. This led to people starting to ‘quote tweet’, adding their own commentary before they retweeted, and eventually Twitter added this as an option to the retweet button.

But this encouraged, as Weatherall calls it, the ‘dunk mechanism’, in which a quote is used to slam or criticise the original tweet as well as amplify it. For some reason, by making the work of sharing within a community seamless and less visible, Twitter created the tools for global shaming and trolling. As Weatherall recalls thinking when he saw Twitter mobs use retweets to attack their targets - “We might just have handed a 4-year-old a loaded weapon.”

Before the rise of social networks, the ‘work’ done by audiences as they discuss popular culture used to be mostly invisible. In a documentary about the making of the cult British comedy show Blackadder, Richard Curtis talked about what it was like to make TV in the 1980s:

‘This was before the days of ratings, I still don’t know how many people watched any episode of Blackadder [the first series]. I remember I used to wander round Shepherd’s Bush, looking in people’s windows, particularly people with basement flats, to see whether or not anyone was watching. I was watching to see if anyone was watching Blackadder One, because one didn’t know whether or not it had been a success or otherwise.’

When I started work as a Commissioning Editor at the UK broadcaster Channel 4 in 2007, we used to have regular Thursday meetings to discuss the performance of our shows and our competitors over the past week. We’d get presentations from the audience research team about ratings, and discuss reviews in the press, but we didn’t have any direct feedback from the actual viewers themselves.

I vividly remember the first time, in 2009, that one of the other Commissioning Editors brought out her laptop to show tweets in response to the broadcast of her show the previous night. She was incredibly excited, as she could see in real time how the audience were discussing her show.

Within a year, Channel 4 had commissioned Seven Days, a show which added this real-time audience interaction into the reality show format. The conversations on Twitter and Facebook were not only visible to the reality show participants, they were used to influence who stayed and who left the show.

I should admit at this point that I was running the multi-platform team at Channel 4 during Seven Days, and the show was a disaster. We built a website that used real-time calls on Facebook data to visualise this conversation, and a coding error combined with a huge rush of audience attention meant that Facebook blocked our calls within the first ten minutes of the launch episode, crashing the show site and creating huge amounts of negative comments on Twitter and Facebook. The show never quite recovered.

But a few years later, another format, Gogglebox, made by the same production company as Seven Days, found a more productive way to show how we discuss and share popular culture. Gogglebox has a rolling cast of ordinary people, filmed on their sofas and armchairs, as they watch TV together. It makes vernacular art out of the jokes, gasps and asides that we all make when we watch TV together with our friends and family.

In the same way that Tumblr created a kind of equality between original content and the work done by communities to share it, Gogglebox presents our everyday discussions about Television on the same level as the content itself. Unlike Seven Days, it doesn’t try to automate these behaviours. Instead it amplifies not just the original work, but the work done by the audience to contextualise, criticise and share it, bringing them both to the same level of visibility.

I think social software designers can learn a lot from Tumblr and Gogglebox. Just because we can track and analyse audience behaviours around content on digital platforms, it doesn’t mean that we should automate them. If we do that, we run the risk of creating tools that amplify the behaviour, but in so doing, erase it at the same time.

The retweet and share buttons on Twitter and Facebook are, in essence, just volume controls. They give us the opportunity to push something to a larger audience with one button, but they don’t require us to explain why we’re doing this. If you only give people the tools to make something louder, eventually all you will hear is the noisy screech of feedback loops.

At the end of the Buzzfeed interview, Weatherall discusses ways in which Twitter could change the Retweet button, either by limiting the audience you can share to (Whatsapp have recently limited message forwarding in an attempt to reduce the impact of misinformation spreading) or by requiring users to actually click on and read a link in a tweet before they can retweet it.

But this feels like the wrong approach - it tries to limit users’ behaviours instead of understanding the importance of their context. I’m not sure if it’s possible to de-toxify the public spaces that Twitter and Facebook have created. By building tools that amplified sharing behaviours, but made invisible the ‘work’ that we do to contextualise our sharing, they prioritised speed over community. They chose to automate the user behaviours they were seeing, without understanding the important role that little moments of friction can play in building communities.

Automating social behaviours is a great strategy if you want to build the biggest and most data-rich advertising network that the world has ever seen. But if you give people the choice of spending their time hanging out in the middle of Times Square, or hanging out with their friends on their sofa, I’ll bet that in the long run, the sofa will always win.

Theo Wangemann (center), phonograph recording in Edison Laboratory Music Room, circa May 1905. NPS image 29-430-003. (Credit: National Park Service)

Who do I think you are?

When I sit down and start typing into a keyboard, I am writing for at least two groups. The first is for myself - I’ve had a bunch of ideas gradually taking shape in my head, and now it’s the frustrating process of getting the words in the right order.

The second is - well, who exactly? I’m writing this as a Substack newsletter, so I have some stats about how many of you will get it (nearly 200 at the time of writing) and how many will read it (about 70% - which is good, but that other 30% kills me). I recognise some of the names on the emails, but by no means all of them. I know some of you signed up from other newsletters, so can guess at your interests or jobs (digital, arts, media - that sort of thing).

But who are you? Or more precisely, who is the *you* in my head as I’m writing this? Do I have a clear idea of who you are, and what will make you interested enough to keep reading? As I edit and rewrite, who is the ideal reader that I am working for? And why have I imagined you in this way?

The word ‘audience’ has the same latin root as ‘audio’ - audentia - and literally means ‘people within hearing distance’. I spoke at a media event in Munich last week, and even though I was speaking in English, I could see the immediate response from the audience - a few nods here and there, a few checking their phones, a good smattering of applause at the end, and (always the most gratifying thing for a speaker) a few people coming up afterwards to ask questions.

But if you can’t see the audience, you can’t get this immediate feedback. In last week’s post I talked about how Arthur C. Nielsen and George Gallup pioneered ways to measure invisible audiences and turn them into data. This week we’re going to look at the other side of the coin - how we turn that data back into (imaginary) people.

When I worked at the BBC in the 2000s, I went to an away-day organised by the Audience Insight team for senior management. This was part of a move to get BBC leaders closer to their audiences, so they had researched a diverse sample of the British public and filmed them answering questions about their hobbies and media habits. We then had to work in groups to think about what BBC content they would like, and took it in turns to feed this back to the group.

Then the Audience Insight team announced a surprise - the people they interviewed were here, in the same room, and we were going to get one of them assigned to each group so we could find out more about their media consumption.

You could almost feel the air being sucked out of the room as a couple of hundred BBC top brass gasped. The audience was actually here? The people who listened, viewed and clicked on our stories? In the same room as us?

We were given a ‘gamer’ for our group - a nice enough guy in his early 20s who very politely answered the questions of the dozen or so media execs on my table. Yes, he did play games for a couple of hours a day. No, they weren’t all about killing things. No, he didn’t play alone - his friends would come round for the night. Yes, he did find time to watch a bit of television as well. Yes, he did prefer gaming to television. No, he didn’t think that was weird.

The reaction in my group was almost as if an alien had landed. I realised that behind the ratings, media execs had all sorts of assumptions about who their audience was, but none of them were anything like these real people, in the room, with us.

Last week, the Columbia Journalism Review published an excellent research project from James G. Robinson on how journalists imagine their readers. Robinson suggests these imaginary audiences fall into four different groups - their peers; family and friends; their sources; and an ‘institutional audience’ - a kind of archetype created and passed on through the history of the publication they were writing for.

The last one is probably the most recognisable. If you think of a publication you know - like Wired, the New York Times, or Vogue - you can quickly summon a picture of who their typical reader might be. Sometimes these imaginary archetypes are based on actual research, and sometimes they accrue over generations. Robinson shares the story of the New York Daily News, who in the 1920s sent a young researcher to the Lower East Side to find out more about their working class audience. The marketing department synthesised his research into an archetype named ‘Sweeney’ who quickly became a shorthand for their editorial direction, and the focus of a trade advertising campaign with the slogan ‘Tell it to Sweeney!’

These institutional audiences become self-reinforcing. They are passed down to new staffers, as they read the journalism of their peers and subtly adjust their writing style for the conventions and imagined audiences of their predecessors. Sometimes these audiences are based on demographics, and sometimes they’re based on the newspaper’s mission or values. But they’re not the only imaginary audience in play - at any one time there are multiple imaginary audiences in the same organisation, including editors, sources, peers, competitors, and the writer’s own friends and family. This last one is particularly influential - Robinson quotes an editor at a Southwestern US paper saying:

“we’re all kind of the same age … [and] an awful lot of those folks are just now having kids. You can test this. Pick up a paper; there are plenty of stories about how to raise 3-year-olds.”

This anecdote will feel familiar to anyone working in the creative industries. We might get incredibly sophisticated data about our audiences and their desires, but a lot of the time we tell the stories that we think people like us will find interesting. This might affect the career choices we make as creators - would you join a publication or broadcaster if your own interests were wildly out of sync with the ‘institutional audience’ of that organisation?

Sometimes, these imaginary archetypes work across whole sectors, not just individual organisations. In the US TV sector, the ‘Nielsen Family’ is shorthand for the entire ratings system. The Nielsen family is an almost mythological beast, with the power to make or break TV shows with their attention. The phrase comes from the fact that TV ratings are measured by ‘households’, with multiple viewers in each household called a ‘family’ - Nielsen’s own website talks about the process of joining their sampled TV panels using the term ‘Nielsen Families’.

This is a very loaded term. It feels like a very traditional, analogue idea of how people watch TV - mom, dad, 2.4 kids, on the sofa watching linear TV at the same time. The phrase ‘Nielsen Family’ has a cosy ring to it, but the process of who gets chosen to be a Nielsen Family is wrapped in mystery, although in reality there’s dozens of accounts of what it feels like to be a Nielsen Family around the web, and Nielsen’s site has instructional videos with interviews from past Nielsen Families.

Criticism of Nielsen ratings has often been about how representative these imaginary families are of actual US audience behaviour. Is the multi-billion dollar TV industry really driven by a few thousand families (each Nielsen family represents around 50,000 US households), some measured by electronic meters, and some filling in diaries to record what they watch? Do they actually measure how people watch TV now, and are diverse or more transient communities (eg student dorms, prisons, etc) under-represented?

It’s hard to overstate how much the archetype of the Nielsen Family has influenced TV commissioning in the last fifty or so years, but this is now starting to change. Netflix famously doesn’t reveal ratings for its shows (with a few exceptions), and their commissioning decisions are based on analysis of its huge amount of viewer data. With every click and view tracked digitally, Netflix has no need for the sampling techniques of Nielsen’s Families.

But I’d imagine there are new archetypes developing within Netflix to describe their audience. Perhaps they are closer to fandom archetypes built around talent and story IP, or audience behaviours rather than demographics. Maybe Nielsen talks about ‘first day series-bingers’, or ‘weekend movie-marathoners’ rather than ‘Nielsen Families’?

Archetypes exist for imaginary audiences in pretty much every creative industry. Even in sectors that are built around digital delivery and data, we still need to think and communicate using archetypes. Perhaps the most influential archetype in digital design for the last few decades - ‘the user’ - is one that seems to run against the idea of archetypes altogether. Despite that, it has become as powerful an archetype in digital design as ‘Nielsen Families’ were for the TV industry.

When I was at BBC New Media in the early 2000s, we used to talk about the ‘former audience’, a term coined around 2003 by US tech writer Dan Gillmor to describe the rise in citizen journalism. It was a provocative phrase, meant to shake traditional BBC execs out of their imaginary archetypes of passive audiences slumped in front of the TV.

At the same time, software designers were using the term ‘user-centred design’ to describe practises that foreground the needs of users in software design, rather than the brand or client commissioning the work. This came out of Don Norman’s hugely influential book ‘The Design of Everyday Things’, that urged designers to think about the experience of a product as well as its aesthetics.

User centred design processes emphasise the role of research into user needs at the beginning, and throughout, the design process. Unlike the commissioning decisions of newsrooms and broadcasters, user-centred design processes aim to route around our hunches about audiences’ needs, insisting instead on data and direct audience research.

Some of my colleagues at the BBC in the early 2000s later set up the ground-breaking Government Digital Service in the UK, that shifted the culture of digital services in government away from the politics and needs of civil service departments towards the needs of the end users. Ben Terret, the Head of Design at GDS, argued strongly against using any archetypes for imaginary users, even the design personas that are often used in the design process:

“Someone had put some persona posters up and I got angry, saying that the whole country were our users, not some neat persona. I ripped a hole in that piece of paper and stuck it in the window.”

There’s a really interesting duality in Terrett’s quote - he rails against the way that personas create archetypes that obscure a more detailed understanding of user needs, but in saying ‘the whole country’ he suggests an even larger archetype. How on earth can you think about the user when your audience is so huge? The hole in his poster shows how difficult this is - in any 10 minute period, the ‘users’ passing through by the window will have included an incredibly diverse range of ages, genders, backgrounds and needs.

The trouble with the ‘user’ as an archetype is that it is singular. User-centred design usually focuses on an individual person’s interactions with a closed system - the archetypical user story structure uses the first person to describe their needs:

“As a <type of user> I want to <achieve a goal> in order to <reason for that goal>”

But user needs exist in a much more complex social system, and user stories can struggle to capture this. We’re starting to see how digital products designed around singular users can meet those individual needs, whilst creating huge issues for broader societal groups. The algorithmic newsfeeds of Twitter, Facebook and Youtube might help us see more content that we like, but they also enable bad actors to quickly spread misinformation. The UK charity DotEveryone has produced resources to help software designers address these consequences, but the reality is that a lot of software designers focus on the needs of an archetypal set of users or personas, not the more complex interactions of a community, society or country.

Perhaps the archetype we should be using for our current era is not the user, but the mob. The word ‘mob’ is a highly charged one, but it comes closer to describing the unpredictable swirling mass of audience activity and algorithmic amplification that underpins our digital culture.

Bill Wasik’s book ‘And Then There’s This’, written in 2010, is a fantastically rich and immersive account of what it feels like to create online mobs. Wasik was an early participant in the contagious media challenges set up by Jonah Peretti when he was at the NYC digital art organisation Eyebeam. Researching how content went ‘viral’ led to him starting his own ‘Mob Project’, emailing invites anonymously to friends inviting them to participate in a happening in a random downtown NYC location. The events quickly spread, and Wasik’s accounts of the spiralling flash mobs perfectly capture the almost visceral experience of seeing something you’ve created taken over by a fast-paced, hyper-connected audience, and what it feels like when you’re no longer in control:

“As the media frenzy over the mobs grew, so did the mobs themselves. For MOB #4, I sent the mob to a shoe store in SoHo […] I was astonished to see the mob assemble: as I marched with one strand streaming down Lafayette, we saw another mounting a pincers movement around Prince Street from the east, pouring in through the glass doors past the agape mounts of the attendants, perhaps three hundred bodies, packing the space and then, once no one else could enter, crowding around the sidewalk, everyone gawking, taking pictures with cameras, calling friends on cell phones (as the instructions for this mob had ordered), each pretending to be a tourist, all feigning awe - an awe I myself truly felt - to be not merely in New York, but so close to the center of something so big.”

Wasik’s book is only slightly coloured by the negatively possibilities of the mob. The experiences he described are from an earlier, pre-lapsarian era of the social web, when hyper-connected mobs could still be imagined as progressive forces, rather than as vectors for trolls and electoral hacks by nation states.

But it is a hugely valuable book for understanding what it feels like to try and design for the mob, to have the mob as your ‘imaginary audience.’ The mob is not a passive reader or isolated user. The mob is a moment in time, an energy that captures and accelerates the actions of a crowd in a way that is powerful, but ultimately outside of its creators’ control. To design for the mob, you need to think about consequences as much as needs.

The archetypes we use to describe our audiences define what kinds of stories are possible, whether you’re writing for the institutional audience of a newspaper, a Nielsen Family, or a user. These archetypes are how we turn data back into people, how the ‘ghosts’ of our invisible audiences become real to us again.

Ideally, they should be able to capture as much diversity as possible - like the hole in the poster on Ben Terrett’s window. But in reality, archetypes end up focusing on what is politically, economically or culturally most important to the people creating them.

In the end, like all metrics, they are mirrors of the organisation as much as they are the audience. It is only rarely, as in the BBC workshop I attended, that we manage to break through the archetypes and see the mob as real, living people instead.

Sometimes, when you’re researching a subject, you can have an ‘Adam Curtis’ moment. It’s the point where you’re deeply entangled in multiple threads of stories and research, and then find something that suddenly seems to snap them into focus.

The problem is, once you’ve noticed the pattern or link that does this, it’s hard to not start seeing it everywhere. It can end up shaping all your research, and you find yourself less interested in threads that don’t fit this pattern. I love Adam Curtis’ films, but sometimes they seem just a bit too, well, neat. His central arguments feel almost too strong. Can the complex stories of politics, culture and society he weaves really be traced back to one or two key characters and events?

So with that caveat, I think I had an Adam Curtis moment last week in my research on the history of audience metrics. I wanted this week to write about two pioneers of audience measurement - Arthur C Nielsen and Archibald Crossley - who developed the first broadcast radio metrics in the 1920s. I’ve written about Nielsen and Crossley before - in fact, this newsletter is named after that essay.

Nielsen is, I think, one of the most influential and under-studied figures of the twentieth century. Starting with his work developing a Food and Drug Index so that American manufacturers could track sales through their increasingly remote and complex distribution chains, he turned the Nielsen company into a metrics and data empire. He once said to his son ‘If you can put a number on it, then you know something’, and his obsession with turning complex systems into numbers means he was the pioneer of Big Data.

Archibald Crossley, who coined the term ‘ratings’, came from a different background in public opinion polling, The battle to measure broadcast radio audiences was essentially about their two different methodologies - the sampled, opinion poll style of Crossley, and the data-driven style of Nielsen.

When Nielsen discovered the Audimeter at MIT, which measured radio audiences via time-stamped marks in a slowly revolving roll of paper tape, he started the process of turning audiences into numbers. Rather than relying on what we say we listen to when interviewed by a researcher, Nielsen had actual data about what we do, driven by each turn of the radio dial. Nielsen’s numbers ended up beating Crossley’s opinion polls.

I thought this battle between Crossley and Nielsen, between numbers and opinions, was the turning point for the story of audience measurement in the early 20th century, and that Nielsen had won. There is a direct line from Nielsen’s Audimeter tracking a hand turning a radio dial, to Facebook’s data centres tracking our every click or tap on our mobile screens. It was Nielsen who ended up dominating the measurement of the cultural empires of broadcasting in the second half of the 20th century, and laid the ground for our current era of Big Data.

But then last week I got a copy of Susan Ohmer’s 2006 book George Gallup in Hollywood. I knew that George Gallup was part of the group of audience measurement pioneers, along with Nielsen and Crossley, in the 1920s and 1930s, but I assumed that he ended up working mainly on political opinion polling, not in media and culture.

Ohmer’s book tells the story of how George Gallup used his political opinion polling techniques for executives at the emerging Hollywood film empires, including RKO Pictures and Walt Disney. Gallup in fact started as a journalist, holding professorships at Columbia and Northwestern, and moved to New York in 1932 to head the research department at the advertising agency Young and Rubicam. He developed a polling technique to help his mother-in-law win office in Iowa, and then set up the grandly titled American Institute of Public Opinion in 1935, the organisation that would eventually become Gallup.

When he was still teaching journalism, Gallup invented new techniques to measure newspaper readers’ attention - not just how audiences bought newspapers, but how they actually read them. One of his insights was that all newspaper audiences across demographics enjoyed, and spent a lot of time, reading the cartoons on the ‘funny pages’. In fact, on average people only read 15 percent of the newspaper, and it wasn’t the news, but the picture pages, sports, and the comics.

Gallup also worked on radio ratings, improving Crossley’s technique of asking listeners to recall what they listened to by having his interviewers knock on people’s doors and ask them what they were listening to right at that moment - a human version of Nielsen’s Audimeter.

He then famously predicted FDR’s win in the 1936 Presidential election, and the Gallup poll became a byword for American public opinion throughout the 20th century. But he was still a journalist at heart - a colleague once suggested his archive of polling results made him a historian, and he replied “No, I am really more a reporter. I report what American citizens think and feel.”

The fascinating aspect of Ohmer’s book is how Hollywood relied on Gallup’s polling technique for decision making. Gallup was asked to poll about the US public’s attitude to Hollywood stars, their reaction to new technologies like Technicolour, and their responses to preview edits of major films like The Wizard of Oz.

If Neislen’s metrics measured the public’s consumption of culture, Gallup’s polls measured their feelings about it.

So this is my Adam Curtis moment - in the stories of Arthur C Nielsen and George Gallup, we have two pioneers of audience measurement. One focused on measuring our behaviour, the other on our feelings and opinions. There are remarkable similarities in their lives - they both created companies that bore their names and became bywords for their industry - the Gallup poll and Nielsen ratings. They were both born at the turn of the 20th Century - Nielsen in 1897 and Gallup in 1901, and both died in the 1980s, just as computers were ushering in the Information Age that there work had laid so much of the groundwork for.

They each represent a different path for audience measurement in the last 100 years, one on behavioural data, and one on audience opinions. Between them, they invented techniques to measure everything we do, and everything we feel. I have a sense that telling the stories of their parallel lives will be a great lens to tell the history of audience metrics. The process of measuring what we do and what we feel were separate for much of the last 100 years, and this separation is perfectly represented in the lives of Gallup and Nielsen.

Arthur C Nielsen died in June 1980, and Gallup in July 1984. Just a couple of months earlier, in May 1984, Mark Zuckerberg was born, the person who would, by connecting the Facebook social graph to the like button, bring the two worlds of Nielsen and Gallup crashing together and turn Facebook into the ultimate store of data on everything we do, and everything we feel.

Image: CBS Program Analyser. Credit: @joshshepperd h/t to Nathan Martin for the link!

This might be the most interesting time to start a newsletter about the history of audience measurement.

And I mean interesting in the spirit of the apocryphal Chinese proverb ‘may you live in interesting times’. Because we might be at the end of a one hundred year journey to measure the invisible audiences at the end of technological broadcast networks.

In the early twentieth century, the invention of radio broadcast networks (and the telephone services that preceded them) created a fascinating new problem for cultural entrepreneurs - how do you measure invisible audiences? If you don’t have a tangible connection with them, through ticket sales, or by being in the same room as them, how can you tell how many people are listening or viewing?

The last 100 years have been a journey to see how to measure ghosts - how to measure the invisible audiences at the end of technological distribution networks. With every decade, these ghosts have come more and more into focus, ending with a the last ten years of social media and digital advertising that has created unimaginable amounts of data about everything we see, read, click and like.

The globe-spanning empires of radio, TV, music, film and digital media have grown by feeding on this data, and developing metrics that have turned this data into cold, hard cash. Or, if you’re in the arts, annual reports, which in turn mean your funders keep giving you cold, hard cash.

If you’ve subscribed to this newsletter, the chances are that you work in one of these industries. At some point, you have had to use this data to work out whether the things you commission or make are actually any good. You probably hated doing this. It involved spreadsheets, or questionnaires, or a presentation full of badly drawn charts. There would have been jargon involved, like MAUs, or conversion funnels, or ratings, or appreciation indexes.

But underneath the numbing surfaces of those spreadsheets and reports there is a strange history, full of ideas, innovators, politics and power. Every new way of distributing culture has had to develop new ways of measuring audiences. Without metrics, they couldn’t convert attention into money, and without money, they couldn’t grow into billion-dollar global empires.

There are some fascinating stories hidden in the history of audience metrics. Let’s go and find them.

Why Apple’s new login means we might have reached Peak Data

It may be ill-advised, but I’m currently re-watching the entire series of Lost with my teenage daughter. I know we’re going to end up frustrated in the end, but we’re only on Season One at the moment, and one of the things I’ve remembered Lost did so brilliantly was jumping across timelines, moving from the back stories of the characters to their present lives on the island with that ominous wooshing noise.

Like Lost, this newsletter is going to jump across timelines. I’ve spent the last decade or so researching the history of audience metrics, and this newsletter was started partly to help me continue this research. In every newsletter, I’m going to try and provide a mix of historical research and commentary on what’s happening right now. Also like Lost, I don’t yet know where this is going to end up. Apologies.

This week, I want to talk about Apple’s new sign-in service, announced at this week’s WWDC. At one level, you can read this as just the latest attempt by the major platforms to create lock-in for their ecosystems (and if you’re interested in exploring this more, you’d be interested in my other project, the Public Media Stack). Apple is going to force every app developer who currently uses Facebook or Google for sign-in to also offer the new Apple service. This is a shot across the bows of their ecosystem rivals, and will potentially deprive Facebook and Google of tons of valuable audience data.

But the really interesting thing is the announcement that they’re developing the service with a ‘minimum viable data’ strategy. The sign in service will create a unique email address for every service you sign in to, making it almost impossible to track users across services - one of the most important goals of digital ad-serving networks. If you want to stop using a service for ever, Apple will delete the link between the one-time email address they created and your real iCloud email address, which will make it impossible for a service to reconnect with you.

This is the latest part of Apple’s ‘Privacy Matters’ campaign, positioning themselves as the anti-Facebook/Google. They’re pushing hard on their pro-privacy policies at the moment, and clearly feel that this gives them a competitive advantage, especially over competitors who are facing anti-trust proceedings in the US driven mainly because of their exploitation of user data.

So is this a sign we’ve reached Peak Data? After one hundred years of slowly measuring audiences through sampled ratings and surveys, then 15 years of exponential growth of ‘big data’, are we finally reaching the point where the tide turns, and ‘just enough’ strategies replace ‘just in case’?

Well, maybe. But only for a privileged few. If there’s one constant in the economics of audience data over the last 100 years, is that we only get free services if we pay for them with our attention. This has been true for commercial radio and television, free newspapers, mobile games and digital content. If we want privacy, we have to pay for it, and not everyone can afford this. Will the right to become a ghost only be for the people with money to buy premium products?

Back in 2004, my innovation team at the BBC produced a number of scenarios about how children might use the internet in 2014. There’s a lot they got right, but the three scenarios all imagined a mixture of more controlled commercial networks in which privacy and security were strong, and deregulated open networks where risks were higher and users were monetised via their data.

This feels like the direction we’re going in now - after a decade of having our data strip-mined by the big ecosystems, we’re going to see the growth of services moving in the opposite direction, offering more privacy and less tracking. But only if you have the means to pay for it. This is worrying, because, as we found in researching those scenarios, it’s often the people who most need security and privacy who can’t afford to pay for it - children, refugees and immigrants, people in low-income households or with precarious support networks.

From a commercial perspective, there is a very real danger that your most financially secure customers will become increasingly invisible to you, and you’ll have to find new ways to develop a relationship with them. But more worrying to me is that people who can’t afford to pay for privacy will be increasingly giving up their data in return for services that they need to get by in our digital age. We might have reached peak data, but perhaps only for the well-off.

Thanks for reading this first newsletter, and I hope you’ve found it useful. It’s ironic that I’m writing this to an audience of ghosts - I recognise some of the 100 or so of you who have signed up for this first newsletter, but I’d love to know more about you all.

Do you think we’re reaching Peak Data? Are we going to see a split in how richer and poorer audiences’ attention is measured? Is there a specific issue or subject related to audience data that you’d like me to talk about? Have you got any of your own insights or research that you’d like to share?

Please get in touch by replying, or commenting on the web version of this newsletter by clicking the button below. And please do forward this email to anyone you think might be interested - I’d love to build a big and vibrant community around this fascinating subject, so please invite anyone else to join us. Thanks!

Matt

View comments

The way we measure audiences is absolutely critical to the billions of dollars that get spent every year making, distributing and monetising media and culture. But who decides how we measure audiences? And how do these metrics end up deciding the kind of culture we get?

I’ve been researching the history of attention metrics for over a decade, and in this newsletter I’m going to tell stories about the fascinating, odd, and hidden story of how metrics are developed. If you work in media or culture, hopefully these stories will give you a deeper understanding of the history of your industry. And a few useful pointers to possible futures…

How To Measure Ghosts is part of Diffusion Network, a new collection of podcasts and newsletters from Storythings.