Press Release

Google Topics: Google’s cookie replacement still doesn’t work

Google has announced a new “privacy” solution for digital advertising called “Topics”.  We discuss how this works and affects rivals below. MOW has today formally filed a submission to the CMA about Google Topics since it appears to circumvent CMA’s intended oversight role as contemplated in Google’s undertakings –  which are still under consideration by the CMA

Topics leaves competitors at a huge disadvantage and lacks key utility.

Topics will provide publishers and advertisers with three topics a browser may be interested in, based on the last three weeks of activity on a browser.

Many customers may still be interested in ads based on last week’s browsing – others may not and may have moved on – and all of us have more than three interests at any one time.

In short, Topics will make effectively tailoring ads and results much less precise, and will put third party competitors at a disadvantage by throttling the data they have access to.

Topics raises some major concerns about regulation.

The Google announcement assumes a world where it blocks cookies from third party domains (third party cookies or ‘TPCs’).

Google’s commitments ought to provide the UK’s Competition and Markets Authority (CMA) effective scrutiny and oversight to review Google’s browser changes – but Topics is not part of Google’s commitments to the CMA.

This means Google is proceeding without CMA oversight to ensure they are neither distorting competition nor imposing unfair terms on users of Google’s business to consumer software. So, Google Topics is being pushed to the market without appropriate oversight.

Topics is a poor replacement for third-party cookies.

Current cookie information supports a variety of uses – including measuring and improving the effectiveness of advertising and assessing whether it is serving users’ interests.

TPCs allow for cross-site data to be passed on to advertisers and publishers, to help get ads in front of the right people.  

Google has presented no evidence that their suggested “interest-based” replacement will adequately support media owner revenues or marketer effectiveness.

They have not even stated that they intend to exclusively use this technology to monetize their own inventory with interest-based advertising.

Google won’t be hamstrung by Topics

Google has suggested that any user data it can collect ought to be its “First party personal data”, using its privileged position in making browser software and operating systems to continue to monitor people’s cross-site digital activity.

James Rosewell, a spokesperson for MOW said: “Google have dropped the mathematical mumbo jumbo and polished the presentation.

“Topics is another flawed solution attempting to unwittingly herd publishers towards Google’s Walled Garden.

“Smart publishers won’t be fooled. The CMA have even more reason to use interim measures as it’s a direct snub to the modified commitments Google proposed to them.”

This seems to be one rule for Google, and another for everyone else, with the Topics rebrand allowing Google to avoid previous attempts at regulation.

Notes for Editors

About the Movement for an Open Web:

The Movement for an Open Web represents a network of businesses, publishers and advertisers that oppose any and all attempts by trillion-dollar companies to unilaterally restrict competition across the open web for their own benefit.

Google’s proposed changes to cookies and user data collection are a prime example of this – and will be bad for consumers and competitors.

If you want to help oppose these kinds of anti-competitive tech practices, consider joining the Movement for an Open Web.

To find out more about membership, or for press and other enquiries, contact

Background on Topics

Topics will provide data to websites to support what Google calls “interest-based advertising”, or the idea that ads online are based on things you’re likely to be interested in.

This system will work by collecting user browser history over a three-week period, and then assigning each browser to a few different ‘topics’.

These topics will then be used to determine what kind of ads might interest users in the future, allowing advertisers to get their ads in front of people that will, in theory, be interested in them.

On the business end, Ad Tech companies, (who are directly competing with Google’s Ad Tech branches) that manage ad placement for businesses selling and buying adverts, will have to rely on data from Topics.

On the user end, this means that people who have just bought a car using a website like CarGurus may well end up greeted by a tide of irrelevant ads for new cars for weeks after their purchase.

Simply, users’ cross-site activity will be put through a system that automatically categorises users and guesses their interests from the websites they visit.

So, if your company is called Autotrader all will be well – Google will know that you are a trader of autos – and place you in a relevant topic group.

But by using site names to automatically categorise, this blunt system could cause problems for sites like MoonPig, ShutterStock or Selfridges, which could end up in entirely the wrong part of the system.

Header image courtesy of Charles Deluvio on Unsplash (Image licensed for use under the Unsplash license)