Publishers got extra time to prepare for the revolution in online advertising, as Google now wants to phase out third-party cookies in late 2023, two years later than originally planned.

The decision was announced in a blog post Thursday by Vinay Goel, privacy engineering director at Chrome, the world’s most popular Web browser owned by Google.

Third-party cookies are the backbone of online display advertising, the US$250 billion industry that has been using them to identify users across the Web, target and retarget them with personalised ads, and measure effectiveness.

Citing privacy concerns of consumers and regulators, Google announced the phaseout plans in 2020 and proposed several new privacy-friendlier technologies as replacements for third-party cookies. 

The proposals, known as the Privacy Sandbox initiative, were supposed to become new global standards for ad targeting by the end of this year. 

“While there’s considerable progress with this initiative, it’s become clear that more time is needed across the ecosystem to get it right,” Goel wrote today.

What caused the delay? 

By late June, only four out of 30 Privacy Sandbox technologies had been ready for testing, so-called origin trials.

Fearing for up to 70% decline in advertising revenue, publishers and ad tech vendors rushed to prepare for the apocalypse. 

Initially, many hoped for alternative identifiers, such as Unified ID 2.0, to preserve individual targeting, but Google abandoned the idea by March and threw the industry in disarray. 

The Privacy Sandbox allows only to target segments of anonymous people grouped by similar Web behaviours (so-called FLOCs). 

Advertising agencies doubted Google’s internal tests that new technologies achieve 95% of conversions per dollar vs. third-party cookies.

Google engaged in talks with advertising industry stakeholders through bodies such as World Wide Web Consortium, Interactive Advertising Bureau, and INMA.

At an INMA meet-up in April, Chetna Bindra, Google's group product manager for user, trust, privacy and transparency, said that following the industry’s feedback the company had, for example, evolved proposals about how remarketing might work.

In June, competition regulators such as the European Commission announced probes into the Privacy Sandbox to prevent Google from abusing its position in the advertising technology and browser markets to enforce the changes that will strengthen Google’s dominance in online advertising. 

Facing criticism from all sides and scrutiny from regulators, Google slowed down. In the Thursday announcement, Vinay Goel declared: “We need to move at a responsible pace.”

“This will allow sufficient time for public discussion on the right solutions, continued engagement with regulators, and for publishers and the advertising industry to migrate their services,” Goel added. “This is important to avoid jeopardising the business models of many Web publishers which support freely available content.”

What does it mean for publishers?

“The good news is that it moves the industry from fire-fighting to more strategic planning,” commented Daniel Knapp, chief economist of IAB Europe, in an e-mail.

Knapp spoke at this week’s INMA Master Class on Smart Data for News Media. In his opinion, the revolution in online advertising is paused but not called off. 

While targeting FLOCs might not be the final result of the Privacy Sandbox, the shift towards privacy is clear:

• Consumers reject tracking: In 2019, according to GlobalWebIndex, 47% of Internet users globally used an ad-blocker preventing ad tech from creating third-party cookies. When Apple forced apps in 2021 to ask users of its new iOS 14.5 for permission for tracking, 85% refused.

• Privacy laws restrict tracking: For example, since 2018, GDPR in Europe has required sites to ask EU visitors for consent to collect personal data, including creating third-party cookies.

• Tech companies abandon tracking: Other browsers such as Apple’s Safari and Mozilla’s Firefox have been blocking third-party cookies by default since 2017 and 2019, respectively.

Google said Thursday the phaseout of third-party cookies in Chrome will proceed in stages:

• Stage 0, Development: Over the next year, Google promised: a) to discuss prototypes of new technologies in forums, such as W3C, b) to test through numerous trials with ad tech, publishers, and advertisers, and c) to launch changes in Chrome only after considering the industry’s feedback.

• Stage 1, Adoption of Privacy Sandbox: Once testing is complete and all technologies are launched in Chrome, sometime late in 2022, publishers and the advertising industry will get nine months to migrate their services.

• Stage 2, Death of third-party cookies: Starting in mid-2023, Chrome will phase out support for third-party cookies over a three-month period finishing in late 2023.

What does the post-cookie future look like?

Facing the uncertainty, it might be worthwhile to consider what we know for sure.

The world after the cookie apocalypse will likely be built on many pillars, differentiated by identifiability and addressability of individuals with ads:

• Targeting logged-in users is most precise. Publishers can obtain consent from users, as they trade access to content and features for registrations or subscriptions. The challenge is that only a small proportion of news Web site users are logged-in or subscribed. 

• Shared identifiers, such as hashed e-mails, promise to offer the scale and flexibility that advertisers seek for matching lists and targeting.Yet the landscape is fragmented; there are more than 50 ID solutions on the market, and publishers cannot pack them all on their pages. There are also concerns whether such identifiers will stand scrutiny of privacy regulators.

• Publishers can build segments based on their first-party user or content data, and offer them in private or open auctions. While this moves targeting from individuals to groups, it offers reach to non logged-in users.

• Targeting to FLOCs or other cohorts of anonymous users grouped by similar behaviours might be a solution adopted by publishers without the chest of first-party data, or used as a replacement for today’s revenue from open exchanges.

• Last but not least, contextual advertising remains an old but effective way to target users based on assumptions about their interests or intent. Many foresee the comeback of advertising networks offering ads on sites of similar contents. And modern contextual targeting can also be very sophisticated, with content classification driven by AI.

In an April survey of 363 INMA members, most respondents said they felt only moderately prepared for the change (55%), while most media businesses remained reliant on third-party cookie technology (85%). 

Only 13% believed they were very well prepared.

The bet on first-party data

In Thursday’s announcement, publishers got a longer runway for adapting to the new era in which the user’s consent will become the most valuable currency.

When asked about preparations, most INMA members said in April they were still strategising around first-party user and content data (68%), but many had already started registering and logging users (61%).

Only 34% have been building their own first-party user or content data platform, and started selling ads based on first-party user or content data.

Some 20% said they will simply use whatever Google builds.

Among the publishers who placed a big bet on first-party data is Norway’s Amedia. Its focus on online subscriptions led a path to the walled garden at a scale competitive to digital platforms. 

In the first quarter of 2021, 80 mostly local brands of Amedia reached 44% adult Norwegians and logged 84% pageviews. It took Amedia two years to reach the threshold of 50% logged-in pageviews and four years to reach 75%. It started its efforts in 2014.

Asked for the success formula, executive vice-president Pål Nedregotten told INMA in May: “If you’re able to deliver journalism that is perceived by a large enough user group as valuable, then you get them to sign up for a subscription. If you consistently deliver that perceived value, they’ll return again and again. If you’re able to extend and improve that value creation from one journalistic area to another, your reading public will grow.”

Still, in order to further expand its walled garden, Amedia teamed up with a competitor — Aller Media — and formed an ID and advertising joint venture Diar, pooling together 1.8 million registered accounts in a 5-million country.

Subscriber profiles are enriched with public records, allowing precise targeting based on age, gender, household size, income, home address, property type, vehicle, and others.

“This precision is greater than offered by the tech platforms and delivers outstanding results,” said Helene Slettemoen, director of insights and strategy at Diar at this week’s INMA Smart Data Master Class. In one campaign, the precise targeting led to an increase of consideration for a brand from 19% to 44% vs. advertising to the general population.

In the coming weeks, Diar is launching a self-serve ad-buying platform for advertisers. It will improve the advertisers’ experience in buying, optimising, and analysing campaigns -- and hopefully address another gap between publishers and the ad tech.

“Publishers’ first-party data is a gold mine, but it has limitations,” warned Daniel Knapp of IAB Europe in an interview with INMA. “Data on content usage covers only a small slice of people’s activities. It says little about intent, shopping decisions and offline behaviour.”

In his opinion, publishers need to bridge the gap between data on content use and purchase behaviours through, as one example, research and data partnerships.

The end game? “Big publishers who amass first-party data may gain the share in the online advertising market, taking over demand from smaller publishers who can’t get the data,” Knapp said.

If you’d like to subscribe to my bi-weekly newsletter, INMA members can do so here.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .