July 2020 – Shifts in the way data is shared threaten Facebook and other rivals in the $330 billion industry.
You may not have noticed, but the web is on the cusp of a revolution. Groups of engineers around the world are beavering away on the biggest upheaval to it in a decade. The shift has the potential to drastically reshape the power dynamics of the internet and the $330 billion digital advertising industry that supports it.
It’s all about cookies. If the web is an unfathomably complex machine driven by billions upon billions of cogs, then cookies are the lubricant that keeps the thing moving. That’s because the web is, for the most part, funded by ads. And the tiny little text files known as cookies determine which ads get shown to whom. For now. Come next year, that’s all going to change.
The overhaul might also give publishers, not least the media industry, a chance to mend their bungled approach to making money in the era of digital consumption. If they get their act together, that is.
Whenever an ad seems to stalk you, cookies are to blame, but their origin in the mid-1990s was innocent enough. Engineers at the then-popular web browser Netscape needed a way for websites to remember what you’d placed in an e-commerce shopping basket, or simply your preferences. They put a small text file on your device noting the data. Then, when you returned to that site, you didn’t have to log in again, and your basket still contained a VHS of Titanic, a Furby doll, and whatever else people bought in the ’90s.
Over time, advertisers realized that such cookies could also be used to track users across the internet. They started piggybacking on websites, with those sites’ consent, to drop “third-party” cookies onto devices, as distinct from the first-party cookies coming from the websites themselves. (You—the user—are technically the second party, but you don’t produce cookies.) The more websites in an advertising network, the more complete a picture they could build of users’ browsing habits and, by inference, their likely purchasing preferences. Knowing that you were interested in the film Titanic, Furbys, and the BusinessWeek website was valuable data.
“It has historically been really easy for third parties with no legitimate agency or claim to the consumer’s data to gain an understanding about the user and then monetize and re-sell that understanding,” says Wil Schobeiri, chief technology officer at MediaMath Inc., a New York-based firm that manages online ad campaigns. “The debate now is,” he says, “who gets to control that and make decisions about the user and the user’s privacy?”
Apple Inc. Chief Executive Officer Tim Cook has identified user privacy as a key selling point for iPhones, iPads, and Macs. In 2017 the company introduced tools for its Safari web browser that made it easier to block third-party cookies. Last year, Mozilla Corp., the maker of Firefox, followed suit. But the biggest upheaval came in January when Google, whose Chrome browser has more users than everyone else combined, announced that it would phase out third-party cookies over the following 18 months. The ad technology industry took a deep breath.
Google was preempting the inevitable. The European Union’s General Data Protection Regulation, introduced in 2018, had already started to unpick the cookie economy by giving citizens more control over their data and letting them opt out of ad-tracking efforts. An array of copycat legislation such as the California Consumer Privacy Act, which went into effect in January, is rolling out around the world. The cookie’s days were numbered before Google’s announcement.
Discussions now focus on what comes next. Every week, members of the World Wide Web Consortium, or W3C, an international standards organization founded by a creator of the web, Tim Berners-Lee, dial into a video call to work out the options. The standards organization has a group focused on improving web advertising that consists of more than 200 members. About 80 join the weekly call: engineers from Google, Facebook, Microsoft, Amazon.com, Apple; publishing giants such as Hearst Communications, Axel Springer, and the British Broadcasting Corp.; advertising agency groups such as WPP; and plenty more besides. Anyone can join for as little as $300 a year. Discussions continue online in a series of open groups on GitHub, the coding repository and forum owned by Microsoft Corp.
The participants are all trying desperately to nudge the tiller in a direction that satisfies their needs. But the captain steering the ship is undoubtedly Google, whose $162 billion in 2019 revenue makes it the gatekeeper for almost half of all global digital ad spending. It has the biggest ad exchange (where brands bid against one another programmatically to show you an ad every time you open a web page); the biggest ad network (virtual real estate where it can place ads on websites); Android, the biggest mobile operating system; and, in Google Search, Maps, YouTube, and Gmail, the most valuable web properties for serving ads. Most significant, Chrome is the biggest web browser, with a 64% market share, according to the web traffic analysis firm StatCounter.
Even without third-party cookies, Google will be able to track users’ activity so long as they’re on Chrome. Whatever it decides, goes. It already vetoed a plan to extend the role of W3C’s web privacy group, Bloomberg News reported last year. So while your data may no longer be traded willy-nilly around the internet, there’s a good chance that Google will still be able to use your browsing history to target you with ads tailored to either your interests or products you’ve already considered buying. Just how it will do that is as yet unclear. Google dubs its latest proposal Turtledove, which stands for Two Uncorrelated Requests, Then Locally-Executed Decision On Victory. It replaced an earlier effort called Private Interest Groups, Including Noise, or Pigin. If you find the ornithological acronyms confusing, you’re not alone; other advertising technology firms have complained about the lack of clarity.
“Chrome is driving this process, and the rest of us are providing feedback,” Alan Chapell, a lawyer for a number of startups, wrote in a July 6 email to fellow participants in the W3C’s advertising business group, or WABG. “Please note that the success or failure of many in the WABG are already contingent upon maintaining a good relationship with Google. As such, there is likely to be some perceived risk to being openly critical of the dominant market player in a public forum. And that is likely to color some of the feedback provided.”
A slew of companies will go out of business after whatever Google decides to implement. Even Facebook Inc. is likely to encounter some difficulties. At the moment, if you see an ad for a pair of sneakers on the Instagram app on your phone, then decide to buy it a week later through the Chrome browser on your laptop, Facebook can tie the events together through third-party cookies, because you’re likely signed into one of its services on the laptop. It estimates that in cases where an ad was converted into a sale within 30 days, just 20% started and ended on the same app and the same device. Severing Facebook’s ability to measure the effectiveness of its ads is a threat to its business. And if that poses a challenge to the world’s second-biggest ad tech player, imagine the peril to the hundreds of smaller companies.
For another group, publishers, it presents a tremendous opportunity. In web parlance, that means anyone producing online content. But it’s particularly acute for the news industry, whose failure to manage the transition from print to digital has been well chronicled: U.S. newspapers’ cumulative advertising revenue fell from a 2006 peak of $49 billion to an estimated $14 billion in 2018, the lowest in 40 years, according to the Pew Research Center. Google’s U.S. ad revenue jumped from $11 billion to $63 billion in the same time frame.
That divergence isn’t just because print advertising has disappeared. It’s also because media organizations have all too willingly outsourced the responsibility for serving their audience with ads to other companies, not least the Google Display Network. That seemed like a good decision 15 years ago, automating the marketing sales process while eliminating the need to invest in costly engineering teams to develop the technology in-house. But it also meant that the value of each ad was tied to the identity of the person looking at it, rather than where it appeared. A brand might pay the same amount for an ad alongside a Pulitzer Prize-winning investigation as it would a fake news story written by a kid in a dorm room, because they were targeting the same person.
That undermined the importance of context, the idea that the website where an ad appears can accentuate or diminish a brand’s appeal. According to Sam Tomlinson, a partner at PwC specializing in the measurement of ad effectiveness, “publishers have always had to have a strong granular understanding of their audience to create the compelling content that sees their audiences coming back time after time.”
The loss of control has been highlighted by Covid-19: U.K. newspapers have warned they could lose £50 million ($63 million) in advertising revenue over three months, even as online readership has spiked. Brands have added vocabulary associated with the coronavirus to so-called blocklists, which means that their ads won’t appear alongside news stories containing those words. While intended to avoid advertising alongside conspiracy theories, the effect has been far broader, and news groups are beholden to the whims of the ad tech industry.
News organizations have in recent years started to make up the ad revenue shortfall by leaning into online subscriptions. And encouraging users to register, whether or not they actually pay a subscription, can also result in harvesting personal information. In the absence of third-party data generated by cookies, the value of first-party data will increase. Knowing someone’s name, address, and interests is valuable: A user can be grouped with a cohort of others with similar interests and income levels—with that person’s permission.
The New York Times Co. has just started that approach. It’s inviting users to fill out an optional questionnaire and informing them that the results will be used to build audience segments that target ads on its website. “We are not interested in sharing our users’ information into broad open networks that re-create the cookie universe we’re in today,” says Allison Murphy, the newspaper’s head of ad innovation. Instead, it may sign agreements with brands that have their own first-party data and target ads that way. “If both of us have first-party data, we can match that in environments that are secure and respect the consent we have from users,” Murphy says.
Given the regulatory pressure against Google, it may also be in the Mountain View, Calif., company’s interest to reinvigorate the media industry’s revenue streams. The U.S. Department of Justice could bring an antitrust case against the company as soon as this summer. Sacrificing a chunk of the $22 billion a year it makes from the Google Display Network might help its antitrust case, even as it squeezes out other ad tech players by making the case for better privacy. That’s surely an outcome Chief Executive Officer Sundar Pichai would prefer to a deeper breakup of Google’s ad tech business and the control it has over the buying, selling, and bidding processes.
If they can get it right, media organizations might be able to reclaim some of the billions of ad dollars hemorrhaged over the past 15 years. Sure, it will be easier for companies with the scale of the New York Times, Condé Nast, or Gannett than it will for independent regional organizations. But the cookies are disappearing from the jar, and everyone should prepare to scramble to catch the crumbs.