Declared that Google will develop a mechanism that 'protects user information while maintaining the relevance of advertisements'
August 22, 2019, Google ’s web browser Chrome development blog says, “A new initiative to support a free and open ecosystem without compromising advertising while improving user privacy protection” Was proposed. Google is seeking input and cooperation from the web community on this proposal, named “ Privacy Sandbox ”.
Building a more private web
Chromium Blog: Potential uses for the Privacy Sandbox
Justin Shu, engineering director at Chrome, pointed out that the technology used by publishers and advertisers to display relevant ads to users is being used beyond its original intent. Originally, data for displaying advertisements suitable for users is collected and used to a level that violates user privacy.
On the other hand, it also states that attempts to improve user privacy have resulted in unintended consequences. For example, by blocking cookies on a large scale to protect user privacy, web developers use browser fingerprints to combine a small amount of information that varies from user to user, such as the devices they own and the fonts they use. Shoe points out that a way to identify users across sites has been created.
In addition, websites that block cookies have been shown to reduce publisher revenue by 52% . Many websites cover the operating expenses with the revenue from advertising, and as a result, online users can enjoy free content. Therefore, if the advertising revenue of content publishers decreases, it seems likely that the content that can be accessed free of charge on the Internet will decrease.
That ’s why Google advocates the “Privacy Sandbox”. The privacy sandbox is a proposal to create a mechanism that protects user privacy and does not impair the relevance of advertisements. It is beneficial to both users and publishers.
Google first argues that it is necessary to identify how user information is used in the advertising ecosystem in order to achieve privacy protection through a “privacy sandbox”.
The next challenge is to build a system that displays advertisements suitable for users without sharing information about user privacy as much as possible. In this regard, Google advocates a way to identify users as large groups with similar trends and preferences and serve ads to each group. By using a machine learning approach called “ Federated learning ”, the browser is a group to which the user belongs until it is confirmed that the group to which the user belongs is a large one that includes thousands of other users. It is said that it is not necessary to disclose.
In addition, Google believes that the criteria for publishers and advertisers when evaluating their ads need to be reviewed. Displaying relevant ads to your users is beneficial not only because advertisers can run effective ads, but also because they do n’t see ads that are n’t of interest to users. An important point for advertisers and publishers is “how the ad provides value that meets the needs of the advertiser” and not “how much personal information the user has collected”. Therefore, it seems necessary to reconsider how to measure the value of ads.
Of course, Google doesn't want advertisers and publishers to suffer from the introduction of a new privacy sandbox. As such, Shu said that Google will take steps to prevent malicious user fraud against advertisers and publishers.
by Soumil Kumar