By Alexandra Bruell
When Google last year said it eventually would stop supporting third-party cookies in its Chrome web browser, and that it would create more privacy-minded ad-targeting tools, marketers expected to still be able to send ads to desirable individual consumers.
But the Alphabet Inc. unit this month said that it also won't provide or use alternatives that uniquely identify people as they move across the internet.
Here are five things we know about the update, according to digital ad executives and people familiar with Google's efforts.
Can advertisers still use their own data to buy targeted ads on Google properties?
On Google properties like YouTube, Gmail and its search engine, yes: Brands will be able to target individuals in certain circumstances. They will need to bring their first-party data into Google through the company's existing Customer Match product, or a similar future product, which ingests information a consumer gave to a brand, such as an email address, and determines whether it matches data Google already has. If so, advertisers can work with Google to send an ad to that person within the company's walls.
Will they be able to use their data outside Google properties?
Outside Google's walls, no: Advertisers no longer will be able to use third-party data to target ads through Google technology, such as the company's automated ad-buying system Display & Video 360, often called DV360. They will need to rely on Google's new FLoC approach. FLoC, which stands for "Federated Learning of Cohorts," will enable targeting ads to groups of people with similar characteristics.
There is one caveat. Brands that have relationships with publishers can pass first-party data directly to those publishers, outside the Google environment, and still manage the buy inside DV360. The use of data and ad buy, however, will be supported by proposed methods that enhance privacy.
What is FLoC exactly, and will it work as well as cookies?
The idea behind FLoC, one of a handful of advertiser tools that Google is proposing, is to allow advertisers to request and use pools of online identities with common characteristics, rather than individuals. The underlying data will come from browsers that use machine learning to develop cohorts based in part on the sites that individuals visit. Advertisers will receive an identifier for a cohort rather than for the individuals within it.
It is hard to know how effective FLoC will be until Google starts testing it with advertisers in the second quarter. But Google said internal tests showed that advertisers in certain instances can expect at least 95% of the desired consumer actions per dollar spent that they get with cookie-based advertising
FLoC is more advanced than Google's other proposed tools and methods, which also includes a proposal called Fledge to enable some level of retargeting -- sending an ad to someone who previously interacted with a brand's website, for example.
Fledge is meant to let an advertiser use an ad server to upload its own first-party data, Google said. Web browsers would then fetch that information, with certain anonymization thresholds, and serve ads to users.
How will measurement work without cookies?
Google is exploring new approaches to measurement, including telling advertisers that specific ads were clicked on, for example, the company said. It is also looking into aggregate reporting that wouldn't associate clicks with specific users, but instead tell advertisers whether a group of people saw an ad or took an action.
"What we're confident of no matter what direction this takes is that we are no longer going to be able to extract data from the ad server log files or from Google at the individual ID level," said John Lee, chief strategy officer at data-marketing firm Merkle.
Without individual-level targeting using cookies, advertisers are also wondering whether they will be able to cap the frequency with which an ad appears to an individual.
When advertisers use first-party data to target consumers on Google properties, they will be able to keep using another company product, Ads Data Hub, for insights about their campaigns.
How are marketers preparing?
Many are focusing on efforts that help them collect more first-party data, such as using first-party cookies to see who visited their sites or running promotions to collect email addresses. As advertisers boost their e-commerce businesses, they also are collecting purchase data.
And many are moving from cookie-based data-management platforms to platforms that manage first-party data.
"What a lot of clients are trying to figure out, if they have a lot of first-party data, is how will it connect in this new world order," said Krystal Olivieri, GroupM's senior vice president for global data strategy and partnerships. "It's a reimagination for how digital will work and in a way that takes consumer rights and choice, but also consumer ethics, into account."
Write to Alexandra Bruell at email@example.com
(END) Dow Jones Newswires