Madmen and the Godless Algorithm

FB-vs-Google

This article from The New Yorker.

Good overview history of the advertising model that has dominated our commercialism for decades. It’s now gone on digital steroids. The disruption of ad technology has interesting implications.

How the Math Men Overthrew the Mad Men

By Ken Auletta

Once, Mad Men ruled advertising. They’ve now been eclipsed by Math Men—the engineers and data scientists whose province is machines, algorithms, pureed data, and artificial intelligence. Yet Math Men are beleaguered, as Mark Zuckerberg demonstrated when he humbled himself before Congress, in April. Math Men’s adoration of data—coupled with their truculence and an arrogant conviction that their “science” is nearly flawless—has aroused government anger, much as Microsoft did two decades ago.

The power of Math Men is awesome. Google and Facebook each has a market value exceeding the combined value of the six largest advertising and marketing holding companies. Together, they claim six out of every ten dollars spent on digital advertising, and nine out of ten new digital ad dollars. They have become more dominant in what is estimated to be an up to two-trillion-dollar annual global advertising and marketing business. Facebook alone generates more ad dollars than all of America’s newspapers, and Google has twice the ad revenues of Facebook.

In the advertising world, Big Data is the Holy Grail, because it enables marketers to target messages to individuals rather than general groups, creating what’s called addressable advertising. And only the digital giants possess state-of-the-art Big Data. “The game is no longer about sending you a mail order catalogue or even about targeting online advertising,” Shoshana Zuboff, a professor of business administration at the Harvard Business School, wrote on faz.net, in 2016. “The game is selling access to the real-time flow of your daily life—your reality—in order to directly influence and modify your behavior for profit.” Success at this “game” flows to those with the “ability to predict the future—specifically the future of behavior,” Zuboff writes. She dubs this “surveillance capitalism.” [I question whether this will really work as anticipated once everybody is hip to the game.]

However, to thrash just Facebook and Google is to miss the larger truth: everyone in advertising strives to eliminate risk by perfecting targeting data.[This is the essence of what we’re doing here – reducing the risk of uncertainty.] Protecting privacy is not foremost among the concerns of marketers; protecting and expanding their business is. The business model adopted by ad agencies and their clients parallels Facebook and Google’s. Each aims to massage data to better identify potential customers. Each aims to influence consumer behavior. To appreciate how alike their aims are, sit in an agency or client marketing meeting and you will hear wails about Facebook and Google’s “walled garden,” their unwillingness to share data on their users. When Facebook or Google counter that they must protect “the privacy” of their users, advertisers cry foul: You’re using the data to target ads we paid for—why won’t you share it, so that we can use it in other ad campaigns? [But who really owns your data? Even if you choose to give it away?]

This preoccupation with Big Data is also revealed by the trend in the advertising-agency business to have the media agency, not the creative Mad Men team, occupy the prime seat in pitches to clients, because it’s the media agency that harvests the data to help advertising clients better aim at potential consumers. Agencies compete to proclaim their own Big Data horde. W.P.P.’s GroupM, the largest media agency, has quietly assembled what it calls its “secret sauce,” a collection of forty thousand personally identifiable attributes it plans to retain on two hundred million adult Americans. Unlike Facebook or Google, GroupM can’t track most of what we do online. To parade their sensitivity to privacy, agencies reassuringly boast that they don’t know the names of people in their data bank. But they do have your I.P. address, which yields abundant information, including where you live. For marketers, the advantage of being able to track online behavior, the former senior GroupM executive Brian Lesser said—a bit hyperbolically, one hopes—is that “we know what you want even before you know you want it.”[That sounds like adman hubris rather than reality.]

Worried that Brian Lesser’s dream will become a nightmare, ProPublica has ferociously chewed on the Big Data privacy menace like a dog with a bone: in its series “Breaking the Black Box,” it wrote, “Facebook has a particularly comprehensive set of dossiers on its more than two billion members. Every time a Facebook member likes a post, tags a photo, updates their favorite movies in their profile, posts a comment about a politician, or changes their relationship status, Facebook logs it . . . When they use Instagram or WhatsApp on their phone, which are both owned by Facebook, they contribute more data to Facebook’s dossier.” Facebook offers advertisers more than thirteen hundred categories for ad targeting, according to ProPublica.

Google, for its part, has merged all the data it collects from its search, YouTube, and other services, and has introduced an About Me page, which includes your date of birth, phone number, where you work, mailing address, education, where you’ve travelled, your nickname, photo, and e-mail address. Amazon knows even more about you. Since it is the world’s largest store and sees what you’ve actually purchased, its data are unrivalled. Amazon reaches beyond what interests you (revealed by a Google search) or what your friends are saying (on Facebook) to what you actually purchase. With Amazon’s Alexa, it has an agent in your home that not only knows what you bought but when you wake up, what you watch, read, listen to, ask for, and eat. And Amazon is aggressively building up its meager ad sales, which gives it an incentive to exploit its data.

Data excite advertisers. Prowling his London office in jeans, Keith Weed, who oversees marketing and communications for Unilever, one of the world’s largest advertisers, described how mobile phones have elevated data as a marketing tool. “When I started in marketing, we were using secondhand data which was three months old,” he said. “Now with the good old mobile, I have individualized data on people. You don’t need to know their names . . . You know their telephone number. You know where they live, because it’s the same location as their PC.” Weed knows what times of the day you usually browse, watch videos, answer e-mail, travel to the office—and what travel routes you take. “From your mobile, I know whether you stay in four-star or two-star hotels, whether you go to train stations or airports. I use these insights along with what you’re browsing on your PC. I know whether you’re interested in horses or holidays in the Caribbean.” By using programmatic computers to buy ads targeting these individuals, he says, Unilever can “create a hundred thousand permutations of the same ad,” as they recently did with a thirty-second TV ad for Axe toiletries aimed at young men in Brazil. The more Keith Weed knows about a consumer, the better he can aim to target a sale.

Engineers and data scientists vacuum data. They see data as virtuous, yielding clues to the mysteries of human behavior, suggesting efficiencies (including eliminating costly middlemen, like agency Mad Men), offering answers that they believe will better serve consumers, because the marketing message is individualized. The more cool things offered, the more clicks, the more page views, the more user engagement. Data yield facts and advance a quest to be more scientific—free of guesses. As Eric Schmidt, then the executive chairman of Google’s parent company, Alphabet, said at the company’s 2017 shareholder meeting, “We start from the principles of science at Google and Alphabet.”

They believe there is nobility in their quest. By offering individualized marketing messages, they are trading something of value in exchange for a consumer’s attention. They also start from the principle, as the TV networks did, that advertising allows their product to be “free.” But, of course, as their audience swells, so does their data. Sandy Parakilas, who was Facebook’s operations manager on its platform team from 2011 to 2012, put it this way in a scathing Op-Ed for the Times, last November: “The more data it has on offer, the more value it creates for advertisers. That means it has no incentive to police the collection or use of that data—except when negative press or regulators are involved.” For the engineers, the privacy issue—like “fake news” and even fraud—was relegated to the nosebleed bleachers. [This fact should be obvious to all of us.]

With a chorus of marketers and citizens and governments now roaring their concern, the limitations of Math Men loom large. Suddenly, governments in the U.S. are almost as alive to privacy dangers as those in Western Europe, confronting Facebook by asking how the political-data company Cambridge Analytica, employed by Donald Trump’s Presidential campaign, was able to snatch personal data from eighty-seven million individual Facebook profiles. Was Facebook blind—or deliberately mute? Why, they are really asking, should we believe in the infallibility of your machines and your willingness to protect our privacy?

Ad agencies and advertisers have long been uneasy not just with the “walled gardens” of Facebook and Google but with their unwillingness to allow an independent company to monitor their results, as Nielsen does for TV and comScore does online. This mistrust escalated in 2016, when it emerged that Facebook and Google charged advertisers for ads that tricked other machines to believe an ad message was seen by humans when it was not. Advertiser confidence in Facebook was further jolted later in 2016, when it was revealed that the Math Men at Facebook overestimated the average time viewers spent watching video by up to eighty per cent. And in 2017, Math Men took another beating when news broke that Google’s YouTube and Facebook’s machines were inserting friendly ads on unfriendly platforms, including racist sites and porn sites. These were ads targeted by keywords, like “Confederacy” or “race”; placing an ad on a history site might locate it on a Nazi-history site.

The credibility of these digital giants was further subverted when Russian trolls proved how easy it was to disseminate “fake news” on social networks. When told that Facebook’s mechanized defenses had failed to screen out disinformation planted on the social network to sabotage Hillary Clinton’s Presidential campaign, Mark Zuckerberg publicly dismissed the assertion as “pretty crazy,” a position he later conceded was wrong.

By the spring of 2018, Facebook had lost control of its narrative. Their original declared mission—to “connect people” and “build a global community”—had been replaced by an implicit new narrative: we connect advertisers to people.[Indeed, connecting people on a global basis for human interaction really doesn’t make a lot of sense. A global gossip network? Unless, of course, you’re trying to monetize it.] It took Facebook and Google about five years before they figured out how to generate revenue, and today roughly ninety-five percent of Facebook’s dollars and almost ninety percent of Google’s comes from advertising. They enjoy abundant riches because they tantalize advertisers with the promise that they can corral potential customers. This is how Facebook lured developers and app makers by offering them a permissive Graph A.P.I., granting them access to the daily habits and the interchange with friends of its users. This Graph A.P.I. is how Cambridge Analytica got its paws on the data of eighty-seven million Americans.

The humiliating furor this news provoked has not subverted the faith among Math Men that their “science” will prevail. They believe advertising will be further transformed by new scientific advances like artificial intelligence that will allow machines to customize ads, marginalizing human creativity. With algorithms creating profiles of individuals, Airbnb’s then chief marketing officer, Jonathan Mildenhall, told me, “brands can engineer without the need for human creativity.” Machines will craft ads, just as machines will drive cars. But the ad community is increasingly mistrustful of the machines, and of Facebook and Google.[As they should be – the value has been over-hyped.] During a presentation at Advertising Week in New York this past September, Keith Weed offered a report to Facebook and Google. He gave them a mere “C” for policing ad fraud, and a harsher “F” for cross-platform transparency, insisting, “We’ve got to see over the walled gardens.”

That mistrust has gone viral. A powerful case for more government regulation of the digital giants was made by The Economist, a classically conservative publication that also endorsed the government’s antitrust prosecution of Microsoft, in 1999. The magazine editorialized, in May, 2017, that governments must better police the five digital giants—Facebook, Google, Amazon, Apple, and Microsoft—because data were “the oil of the digital era”: “Old ways of thinking about competition, devised in the era of oil, look outdated in what has come to be called the ‘data economy.’ ” Inevitably, an abundance of data alters the nature of competition, allowing companies to benefit from network effects, with users multiplying and companies amassing wealth to swallow potential competitors.

The politics of Silicon Valley is left of center, but its disdain for government regulation has been right of center. This is changing. A Who’s Who of Silicon notables—Tim Berners-Lee, Tim Cook, Ev Williams, Sean Parker, and Tony Fadell, among others—have harshly criticized the social harm imposed by the digital giants. Marc Benioff, the C.E.O. of Salesforce.com—echoing similar sentiments expressed by Berners-Lee—has said, “The government is going to have to be involved. You do it exactly the same way you regulated the cigarette industry.”

Cries for regulating the digital giants are almost as loud today as they were to break up Microsoft in the late nineties. Congress insisted that Facebook’s Zuckerberg, not his minions, testify. The Federal Trade Commission is investigating Facebook’s manipulation of user data. Thirty-seven state attorneys general have joined a demand to learn how Facebook safeguards privacy. The European Union has imposed huge fines on Google and wants to inspect Google’s crown jewels—its search algorithms—claiming that Google’s search results are skewed to favor their own sites. The E.U.’s twenty-eight countries this May imposed a General Data Protection Regulation to protect the privacy of users, requiring that citizens must choose to opt in before companies can horde their data.

Here’s where advertisers and the digital giants lock arms: they speak with one voice in opposing opt-in legislation, which would deny access to data without the permission of users. If consumers wish to deny advertisers access to their cookies—their data—they agree: the consumer must voluntarily opt out, meaning they must endure a cumbersome and confusing series of online steps. Amid the furor about Facebook and Google, remember these twinned and rarely acknowledged truisms: more data probably equals less privacy, while more privacy equals less advertising revenue. Thus, those who rely on advertising have business reasons to remain tone-deaf to privacy concerns.

Those reliant on advertising know: the disruption that earlier slammed the music, newspaper, magazine, taxi, and retail industries now upends advertising. Agencies are being challenged by a host of competitive frenemies: by consulting and public-relations companies that have jumped into their business; by platform customers like Google and Facebook but also the Times, NBC, and Buzzfeed, that now double as ad agencies and talk directly to their clients; by clients that increasingly perform advertising functions in-house.

But the foremost frenemy is the public, which poses an existential threat not just to agencies but to Facebook and the ad revenues on which most media rely. Citizens protest annoying, interruptive advertising, particularly on their mobile phones—a device as personal as a purse or wallet. An estimated twenty per cent of Americans, and one-third of Western Europeans, employ ad-blocker software. More than half of those who record programs on their DVRs choose to skip the ads. Netflix and Amazon, among others, have accustomed viewers to watch what they want when they want, without commercial interruption.

Understandably, those dependent on ad dollars quake. The advertising and marketing world scrambles for new ways to reach consumers. Big Data, they believe, promises ways they might better communicate with annoyed consumers—maybe unlock ways that ads can be embraced as a useful individual service rather than as an interruption. If Big Data’s use is circumscribed to protect privacy, the advertising business will suffer. In this core conviction, at least, Mad Men and Math Men are alike.

This piece is partially adapted from Auletta’s forthcoming book, “Frenemies: The Epic Disruption of the Ad Business (and Everything Else).”

 

I would guess that the ad business will be disrupted further as we find new ways to connect consumers with what they want. This will reduce the power of the Math Men at centralized network servers.

I also suspect search will become a regulated public utility. A free society cannot tolerate one or two private corporations controlling all the information data that flows through its networks.

 

Why musicians are so angry…

…at the world’s most popular music streaming service

Slide09
Washington Post
 July 14
With the money from CDs and digital downloads disappearing, the music industry has pinned its hope for the future on online song streaming, which now accounts for the majority of the $7.7 billion U.S. music market.

But the biggest player in this future isn’t one of the names most associated with streaming — Spotify, Amazon, Pandora or Apple. It’s YouTube, the site best known for viral videos, which accounts for 25 percent of all music streamed worldwide, far more than any other site.

Now, YouTube is locked in an increasingly bitter battle with music labels over how much it pays to stream their songs — and at stake is not just the finances of the music industry but also the way that millions of people around the world have grown accustomed to listening to music: free of cost.

Music labels accuse YouTube of using a legal loophole to pay less for songs than traditional music-streaming sites, calling YouTube the biggest threat since song piracy crippled the industry in the early 2000s. The industry has pressed its case to regulators around the world in hopes of forcing a change.

“I do think YouTube is starting to panic a little bit,” said Mitch Glazier, president of the Recording Industry Association of America.

But YouTube is not backing down, stressing the benefits to musicians of promotion on one of the Web’s most popular sites — which allows ordinary users to integrate music into their uploads. YouTube also warns against attacks that could reduce competition among streaming services.

“The industry should be really, really careful because they could close their eyes and wake up with their revenue really concentrated in two, three sources,” said Lyor Cohen, YouTube’s global head of music, referring to Spotify, Apple Music and Amazon Prime Music. (Amazon founder Jeffrey P. Bezos owns The Washington Post.)

The music industry counters they are backed into a corner when negotiating with YouTube — a unit of Google-parent Alphabet — which is mostly shielded by federal law from being responsible for what users post on the site.

“It isn’t a level playing field,” said one executive at a major music label who spoke on the condition of anonymity because he wasn’t authorized to talk, “because ultimately you’re negotiating with a party who is going to have your content no matter what.”

Now, the battle is heating up as the European Union is expected to release new rules later this year for how services such as YouTube handle music, potentially upending some of the copyright protections that undergird the Internet.

Online streaming works like a digital jukebox, with fractions of a penny paid each time a song is played. The money comes from ads and subscriptions.

The E.U. has formally recognized that there is a “value gap” between song royalties and what user-upload services such as YouTube earn from selling ads while playing music. YouTube is by far the largest user-upload site.

How such a law would address the gap is still being decided, but the E.U. has indicated it plans to focus on ensuring copyright holders are “properly remunerated.”

Even the value gap’s existence is disputed.

A recent economic study commissioned by YouTube found no value gap — in fact, the report said YouTube promotes the music industry, and if YouTube stopped playing music, 85 percent of users would flock to services that offered lower or no royalties.

A different study by an independent consulting group pegged the YouTube value gap at more than $650 million in the United States alone.

“YouTube is viewed as a giant obstacle in the path to success for the streaming marketplace,” Glazier said.

The dispute boils down to what YouTube pays for songs.

Musicians from Arcade Fire to Garth Brooks to Pharrell Williams say they earn significantly less when their songs are played on YouTube than on a site such as Spotify — even though many listeners use these services in the same way. Both YouTube and Spotify allow users to search for music and find song recommendations. On YouTube, users can find music alongside cat videos and toy reviews in what is generally a free-for-all of content, while people go to Spotify and the like for a more refined experience. Some audiophiles argue the sound quality on music streaming sites is superior.

YouTube pays an estimated $1 per 1,000 plays on average, while Spotify and Apple music pay a rate closer to $7.

Irving Azoff, the legendary manager for acts such as the Eagles and Christina Aguilera, said he has one artist — whom he declined to name — who gets 33 percent of her online streams from YouTube but only 10 percent of her streaming revenue.

Smaller acts see it, too. Zoe Keating, an instrumental cello player, showed The Washington Post a statement from YouTube showing that she earned $261 from 1.42 million views on YouTube. In comparison, she earned $940 from 230,000 streams on Spotify.

“YouTube revenue is so negligible that I stopped paying attention to it,” Keating said.

YouTube admits that it pays less for songs.

But the reason for this disparity is where the two sides split.

The music industry claims YouTube has avoided paying a fair-market rate by hiding behind broad legal protections. In the United States, that’s the “safe harbor” provision, which essentially says YouTube is not to blame if someone uploads a copy-protected song —unless the copyright holder complains.

This, the music industry argues, leads to a costly game of “Whac-A-Mole”: hunting for illicit song uploads and filing notices with YouTube.

“You can’t prevent something from going up on YouTube. All you can do is ask them to take it down,” said Stephen Carlisle, who runs the copyright office at Nova Southeastern University. “At some point, it’s not worth it to do this.”

YouTube says it has the solution: Its Content ID system automatically checks for violations by comparing songs detected in new uploads against a database of claimed songs, capturing 99.5 percent of complaints. The company says it averages fewer than 1,500 traditional copyright claims from the music industry a week.

YouTube also pointed out that it has licensing deals with music labels large and small.

Earlier this year, Warner Music Group — one of the “big three” music labels — signed a new licensing deal with YouTube, and a memo from Warner chief executive Steve Cooper leaked out, saying the deal was signed “under very difficult circumstances.”

“There’s no getting around the fact that, even if YouTube doesn’t have licenses, our music will still be available but not monetized at all,” the memo continued.

Warner confirmed the memo’s authenticity, but, like the other major labels, declined to comment for this article.

Cooper’s complaints surprised Cohen, who worked at Warner until leaving for YouTube last year.

“I never heard that from his mouth during the entire negotiation,” Cohen said.

Cohen’s move to YouTube created waves in the industry. After all, Cohen was famous for taking one of the hardest stands against YouTube when, in 2008, he pulled Warner’s entire song catalogue from the video service to protest low song royalties. It was the nuclear option.

And it failed. After nine months and spending $2 million trying to keep its music off YouTube, Warner capitulated.

Cohen said he was sympathetic to his former colleague’s complaints. But YouTube pays $1 billion in song royalties worldwide each year. Cohen said his company has been hindered by its global reach — ad rates are lower outside the United States — and its slower rollout of a subscription option, YouTube Red. Song royalties are higher with monthly subscriptions than ads.

“What I’m trying to do with YouTube is be a cheerleader to build a subscription business that the industry can be proud of,” Cohen said.

Nabila Hisham, 22, is a music fan on YouTube. Recently, the college student in Kuala Lumpur, Malaysia, has been playing one song repeatedly: “Despacito,” a chart-topping Latin pop remix featuring Justin Bieber. The YouTube video — which has a total of 412 million plays — is a photo of Bieber’s tattooed neck. The video is beside the point. For, Hisham, it’s about the music.

“I’m glad that YouTube exists,” she said.

Correction: A previous version of this story stated YouTube’s ContentID system automatically handles 98 percent of copyright management for songs. The system handles 99.5 percent.