The Federated Learning of Cohorts (FLoC) technology, which will replace cookies, was announced by Google last year. The company recently announced a technology launch postponement to 2023, and now the digital marketing market has less than a year to find an alternative to working with data.
How cookies work
Cookies work in the following way: the user visits a site where partner pixels are installed. Each such pixel sets its own 3rd party cookie, which contains an impersonal user ID. Through her, he reports that the browser, for example, is used by a 32-year-old woman who is interested in yoga, traveling, drives a mini-cooper, and wants to take out a mortgage.
Advertisers use DMP to show the user current offers — yoga studios, relevant residential complexes, or mortgage offers — advertisers use DMP. This is a platform that allows you to collect, process, and track visitor actions. In this case, data management is reduced to the maximum collection of information from all possible partner sites through cookies. After that, the advertiser is provided with data that can be uploaded to advertising accounts. This is how the best personalization is achieved.
In addition to the fact that cookies are tied to a specific user, the data, although encrypted, migrates through a large number of private players. Google plans to create its own segments that will not leave the user’s browser.
Why FLOC is needed
The main task of the technology is to protect the personal data of users and maximize privacy. With it, a specific user will not be identified by the cookies that his browser gives. The new technology will continue to manage data, but not in the focus of a specific user, but a group of people with a similar behavioral pattern. FLoC is expected to offer advertisers updated user segments.
Their interests will be determined using short labels – cohorts, which will be calculated on the browser side. The details will remain on the user’s side, and only the cohort ID will be given outside. This will allow you to serve relevant ads without tracking a specific person.
Alternative targeting methods are already operating in the Western market. At the moment, about a dozen traffic identifiers have been created, for example, Unified Data, ID5, and others. The most popular are the following:
- Authenticated traffic. Data collection is due to e-mails and phones
- Fingerprinting. This is a unique, impersonal fingerprint of the system, the identifiers of which can reveal characteristics about the owner. However, traditional tracking methods such as IP addresses and cookies are not used to collect data.
- Hybrid option. It includes solutions of both the first method and the second.
In addition to fingerprinting, ready-made solutions offered by the market have several disadvantages:
- Small overlap of audiences. This is due to the fact that the patterns are collected from the traffic of foreign users.
- Personal data authentication is applicable to the e-commerce market. This sector uses numbers and e-mail to a greater extent to enter the site, and this does not solve the problem on a global scale.
- Use of a number of identifiers on a commercial basis. ID5 technologies and others can only be used on a commercial basis, which requires certain resources from the company.
Now companies should focus on collecting information about users: collect offline user IDs, match them with cookies through CDP / DMP systems, record website visitors, choose a DMP platform, and increase 1st party data – for example, by creating a personal account on the site.