Google has launched a new proposal for a successor to cookies from its Privacy Sandbox initiative, Topics, which replaces its previous proposal, Federated Learning of Cohorts (FLoC).
Topics lets a web browser determine a number of interests based on a user’s browsing history over the previous three weeks, with older information deleted. When a user visits a site using Topics, the technology selects three topics, one from each of the previous three weeks, to share for advertising purposes.
In Chrome, Google plans to provide controls that will allow users to see their topics, remove individual topics, and disable the feature altogether.
Google said that Topics was intended to support similar use cases to FLoC, but had significantly different design and functionality. It does not involve the concept of "cohorts" of users who share common behaviours and, unlike in FLoC, Topics will not make use of sensitive categories such as race or gender. Other differences from FLoC are outlined in a technical explainer.
James Parker, chief solutions officer for data and planning at Jellyfish, welcomed Google's announcement, while acknowledging it would require a change to the approach used by many brands and agencies.
"Firstly, it's great to see that Google is listening and adjusting to consumer and industry feedback," he said. "It's not the news that some advertisers will want to hear, but this is a step in the right direction for our industry. It will require a refocus on how marketing and agency teams are now organised, as there are a lot of people in the industry who only know how to buy digital media in a certain way and they will need to be retrained."
Advertisers, Parker said, would likely "react by doubling down on contextual buys and organised plans publisher by publisher. This will impact revenues and potentially limit advertiser reach and increase the prices of the major publishers."
Noting that "its been very quiet in terms of news recently" from Sandbox, Mark Syal, global chief product officer at Brainlabs, acknowledged that Topics "may feel less ambitious than FLoC" in what it can offer to advertisers.
"It has not been easy to work out how to target advertisers where we have no direct first-party connection and target them with meaningful ads while maintaining their privacy and control over their own data," he said. "But I think it's a necessary step if we want to create a sustainable advertising ecosystem.
"I would rather have something workable that produces good quality targeting options than something that will be phased out eventually."
Google’s methods for enabling effectively targeted ads in the post-cookie world are competing with rival technologies such as Unified ID 2.0, which is led by The Trade Desk and supported by a raft of other ad tech companies. UID 2.0 uses hashed email addresses for advertisers to target individual users.
Last March, Google said that following the demise of cookies, it would not allow technologies such as UID 2.0 that track individual users across its own products, which include buying platform DV 360, publisher platform Google Ad Manager and Adsense.
While this policy will not stop publishers and advertisers using such tools elsewhere, Tom Kershaw – the former chief technology officer of Magnite and chairman of Prebid.org, the organisation operating UID 2.0 – told Campaign Asia last year that that announcement was “a statement in favour of the Privacy Sandbox and against the user login initiative that the industry is pursuing”, but would not actually impact UID 2.0.
Phil Duffield, UK vice-president at The Trade Desk, said he supported Google's efforts to find privacy friendly alternatives to cookies, but warned that Topics had limitations.
"It is important to note that Google’s latest proposal on Topics is limited to one channel [its Chrome browser] and can’t be applied to the multiple devices the majority of consumers use to access the open internet," he said. "This means that advertisers are restricted to basic targeting approaches, which limits their full potential.
"Users, in turn, also receive a suboptimal experience online. Additional iteration will be required to address key advertising use cases and adequately answer the question of how to support the fundamental value exchange of the internet."
In contrast, he called UID 2.0 "an industry-wide identity solution, a new approach to identity that reflects the contemporary digital marketplace".