In the latest update on its plans to replace third-party cookies for advertising, Google said testing on a particular proposal looks promising.
Google planned to share some new findings that show the effectiveness of its “Federated Cohort Learning” proposal that is part of the Chrome browser’s “Privacy Sandbox” in a blog published on Monday. The “Sandbox” is an initiative launched in 2019 to find alternatives to the cookie and at the same time mitigate the impact on publishers and other players. In Google’s words, it was about finding a solution that protects user privacy and allows content to be available for free on the open web.
Shortly after announcing the initiative, Google said it would end support for third-party cookies, which feed much of the digital advertising ecosystem, on its Chrome browser within two years of January 2020.
Chrome engineers have been working with the broader industry, including organizing W3C web standards, on ideas in the Sandbox that were proposed by Google and other ad technology players. What will likely result is a series of these ideas moving forward, says Google.
“This is a proposal,” Chetna Bindra, product group manager for user trust and privacy at Google, told CNBC about the progress of “FLoC”. “It is not at all the final or singular proposal to replace third party cookies … There will be no final API that will go ahead, it will be a collection of them that allows things like interest-based advertising, as well as for metric use cases, where it is critical to being able to ensure that advertisers can measure the effectiveness of their ads. ”
Bindra said the company is “extremely confident” about the progress of the proposals and tests so far.
The Google post on Monday says the test results show that FLoC (pronounced like a flock of birds, according to a series of proposals with bird themes like “Turtledove” and “Swallow”) is “a sign of effective privacy-focused replacement for third-party cookies. “He says advertisers can expect to see at least 95% of conversions per dollar spent compared to cookie-based advertising.
FLoC would basically put people in groups based on similar browsing behaviors, which means that only “cohort IDs” and not individual user IDs would be used to target them. The web history and entries for the algorithm would be kept in the browser, with the browser only exposing a “cohort” that holds thousands of people.
“We are really seeing that one of those early Sandbox technologies for interest-based ads is literally almost as effective as third-party cookies,” said Bindra. “There are certainly a lot more tests coming up. We are very excited for advertisers and ad technology to get involved directly.”
Bindra said that these cohorts, which may include people who have behaviors like an interest in gardening or rock music, would still allow targets based on those interests. Rather than focusing on an individual level, however, it would target groups.
“The difference really will be that now they are no longer tracking all users of the web. There really is this notion of privacy for those users who are now grouped into a group,” said Bindra.
She added that FLoC test numbers should be reassuring for publishers. Next, Chrome will make the cohorts available for public testing at its next launch in March, and expects to start testing FLoC-based cohorts with advertisers on Google Ads in the second quarter, says the blog post.
Myles Younger, senior director of global data practice at MightyHive, said that all Sandbox proposals are coming to “how can we build new features in the Chrome browser to simultaneously resolve user privacy and the death of a third-party cookie while preserving capacity of brands to advertise effectively. “He spoke before the latest Google findings were released.
One question is whether players will actually use it.
“I’m not sure if it’s something that Google is able to just flip a button on and turn it on,” he said. “Publishers have to use it. People have to start using this system. [Google] you need to prove it works. ”
Paul Bannister, director of strategy for CafeMedia, said that advertisers and publishers have some fear of the unknown when it comes to what comes next.
“I think we all want to believe that this is going to be good and we all want to get to a place where users have more privacy and the web works better,” he said. But considering how complicated and technical the process is, it is unclear what will really happen next.
He said there was some fear that such action could benefit the “walled gardens” of companies like Facebook, and far from open web advertising.
UK antitrust authorities are also keeping an eye on the plans and are investigating whether the plan to remove third-party cookies from Chrome could harm online ad competition. The Competition and Markets Authority is investigating whether Google’s plans can lead advertisers to direct spending to Google’s own tools at the expense of its competitors.
In an email response, Bindra said, “Privacy Sandbox has been an open initiative from the beginning and we appreciate the CMA’s involvement as we work to develop new proposals to support a healthy web supported by third-party cookie-free ads.”
Some privacy advocates are also skeptical of the “FLoC” approach. The Electronic Frontier Foundation wrote in 2019 that these cohorts can be used in harmful ways, allowing discriminatory advertisers to identify and filter groups that represent vulnerable populations.
“A herd name would essentially be a behavioral credit score: a tattoo on your digital forehead that provides a succinct summary of who you are, what you like, where you go, what you buy and who you associate with,” a technologist on the team at EFF Bennett Cyphers wrote in the blog post. “The herd names are likely to be inscrutable to users, but they can reveal extremely confidential information to third parties.”
Whether machine learning would create cohorts based on health problems or low income status or other confidential attributes is a question for some.
“He can do potentially scary and possibly illegal things,” said Bannister. “How will Chrome protect itself from this?”
Google said in documents that its analysis assesses whether a cohort can be sensitive without learning why it is sensitive, and said that cohorts that reveal “sensitive categories” such as race, sexuality or personal difficulties have been blocked or the grouping algorithms have been reconfigured to reduce the correlation.
Google added that it is against its policies to serve personalized ads in these sensitive categories.
Applications are open for 2021 CNBC Disruptor 50, a list of private start-ups that use innovative technology to become the next generation of large public companies. Submit on Friday, February 12, at 3 pm EST.