In its search for a cookieless advertising ecosystem Google changes Floc cohorts for individual Topics and says the company will unveil another api being currently developped inside its Privacy sandbox called Fledge soon.
Fledge is a solution for remarketing use cases, designed so it cannot be used by third parties to track user browsing behaviour across sites.
Google has decided to stop its FLoC project, changing the focus from big non personalized data or cohorts for a focus on the individual browsing activities of each user through the Topics api.
The end of ad effectiveness and meassuring in Google’s cookieless world?
What Google Topics makes is using your browsing history from the last three weeks to identify your interests online. Initially, some 350 topics (taxonomy) will be analyzed, but the intention is to increase that number. And the new appi just keeps the topics information available during three weeks, after this time period new topics will be generated based on the user’s browsing activity and old ones will be deleted. The api will select a maximum of five topics for each user and three weeks period.
FLoC was Chrome’s initial privacy-preserving proposal to allow companies to show relevant ads to people based on their web browsing activity, without revealing specific sites they visited to any external parties including Google.
Chrome experimented with FLoC in 2021 and with the feedback received from regulators, privacy advocates, developers and industry, it was conveived the new Topics api proposal, which addresses the same general use case as FLoC, but takes a different approach. While FloC was intended to group people with similar browsing histories, Topics seems to offer a individual, rather than group segmented approach for ad targeting.
Chrome intends to experiment with the Topics api worldwide, including Europe, in 2022 and is no longer developing FLoC.
Compared with the FLoC’s identifier generation system topics provides a far less refined information on the user who will receive the ads related to her/him interests (topics).
The coarseness of the topics probably will provide a better privacy protection for Internet surfers than third party cookies and FLoC api, but will surely make the Topics api far less interesting than FLoC for advertisers and researchers. Topics slightly reduces the amount of cross-site identifiable information of each Intenet surfer that advertisers and researchers can obtain with the app.
Fledge hasn’t hit Beta yet, I will wait to comment anythin on this api.
The six reasons why Google changed FLoC for Topics
Google explains there are six reasons why they have decided to leave aside the FLoC project and chaning to the individual Topics approach:
1.- FLoC didn’t actually use Federated learning, so why was it named Federated Learning of Cohorts?: Google wanted to integrate federated learning into FLOC but they found that on-device computation offered enough utility and better privacy.
2.- FLoC added too much fingerprinting data to the ecosystem: the Topics API significantly reduces the amount of cross-site identifiable information. The coarseness of the topics makes each topic a very weak signal; different sites receiving different topics further dilutes its utility for fingerprinting.
3.- Stakeholders wanted the API to provide more user transparency: the Topics API uses a human-readable taxonomy which allows users to recognize which topics are being sent (e.g., in UX). This makes the system more understandable for users.
4.- Stakeholders wanted the api to provide more user controls: with a topic taxonomy, browsers can offer a way (though browser UX may vary) for users to control which topics they want to include. The Topics api in Chome will include controls for the user to delete the topics he/she do not want to share and even swith off the api.
The Topics api will have a user opt-out mechanism, both for websites and people using their browsers.
5.- FLoC cohorts might be sensitive, while Topics api taxonomy excludes sensitive topics: FLoC cohorts had unknown meaning. The Topics API, unlike FLoC, exposes a curated list of topics that are chosen to avoid sensitive topics like race, religion, etc…
It may be possible that topics, or groups of topics, are statistically correlatable with sensitive categories. Google aknowledges that tis is not ideal, but underscores that it’s a statistical inference and provides considerably less personal information on the user than what can be learned from cookies (e.g., cross-site user identifier and full-context of the visited sites which includes the full url and the contents of the pages).
6.- FLoC automatically included browsing activity from sites with ads on them, but stakeholders feed back showed it shouldn’t: To be eligible for generating users’ topics thrugh the Topics api, both, sites and people will have to use the it.
First, Chrome determines the topics associated with websites. For example, a yoga website might be classified as being related to “Fitness”. These topics are assigned to the websites themselves, not to the person surfing the website. Then, Chrome keeps a handful of the surfer’s most frequently visited topics in a given week, and participating sites show this person ads based on these topics.
The calculation of the most frequent topics occurs entirely within the browser, without sharing data with external servers. The resulting topics may be sent to servers in an anonymous way for sensitivity and abuse analysis.
Only sites that are using Topics or have at least one embedded service that uses the Topics API, will be included in topics generation. Sites can also specifically opt-out using Permission Policy, an existing browser mechanism that is used to control features like camera access. If a website had opted out of FLoC, the Topics API will respect that decision.
On the side of the person surfing the Internet, he/she will be able to delete in an easy way the topics to be shown one by one or even switch off the app, so that no topic will be shown to the websites or advertisers using the Topics api.
Topics, not completely secure, but far better than the third parties cookie system for people surfing the Internet, Google says
Google considesr the API to be a step toward improved user privacy on the web, but they warn that Topics api is far from perfect in this regard:
1.- It is possible for an entity (or entities) to cooperate across hosts and acquire up to 15 topics per epoch for the same user in the first week.
2.- There are some risks that other data on the users browsing history appart from the topics can be obtained through the Topics api:
a) It is theoretically possible that a number of different callers that call the API on different sets of sites collude to determine more detail about the sites a user visited, or to accumulate a user identifier over time.i
This is something that the browser could potentially observe and may intervene on if necessary.
b) There is one piece of information that the API reveals that goes beyond the capabilities of third-party cookies: that the topic returned is one of the top 5 browsing topics for the user for the given week.
3.- There are means by which sensitive information of the user may be revealed, despite no sensitive topic will be included in the app taxonomy: the list of topics may have unintended correlations to sensitive topics.
But in the end, Google underscores, “these sensitive information is probabilistic and far less detailed than what cookies can provide from full page content, full urls, and precise cross-site identifiers”.
Image over the headline.- Cookies vs Topics.© Google
Related Eastwind Marketing links (Spanish edition):
Related external links:
To find more technical information on Topics app, click here
To find answer fot themost frequent asked questions, on the Topics app, click here
To find more information on Fledge, click here
