Google shook the foundations of the internet when it announced in 2019 its intention to remove cookies from third parties. Other browsers, such as Firefox or Safari, had already done it before. But it turns out that Chrome is the portal in which around 70% of Internet users browse and that Google is the search engine preferred by the vast majority (87%). If they both stop sharing cookies to its users, in practice it is as if they ceased to exist.
The industry has been plotting how to organize the internet since 2019 when Google’s promise is consumed, which will soon happen in 2023. Facebook, the great competitor of the Mountain View giant, the company that together with Google dominates digital advertising with an iron fist (they control 60% of the market), it is clear that the future depends on cooperation between stakeholders. This is assured by its executives, who appeal to the union of the sector (clients, advertisers and other platforms) to design a satisfactory solution for all parties. “This is not something Facebook can do alone. We have to work together as an industry, ”said Ben Savage, Senior Engineer at software of the company, during an international meeting with journalists.
Dependence on ‘cookies’
Third party cookies are pieces of code that remain on our devices every time we visit a website. The first part, the ones that the same website that we access puts, are necessary for its operation: they allow, for example, that they save our passwords and preferences for the next time we enter. The third-party ones, the ones Google wants to eliminate, are the most valuable to the advertising industry. These cookies They are generated by providers other than the page on which you are browsing. They are used so that a website can know what others you have been on before reaching it, for how long, from what type of device you have entered, if you have bought something or what has interested you the most. They also allow companies to obtain very valuable information from users, such as address, gender or age.
Digital advertising today is structured around the processing of this data, which allows a very deep analysis of who enters a page and for what purpose. Third-party cookies are the essential gear for personalized ads to reach the desired audiences. Every time someone enters Facebook, for example, they take a cookie with them. And, at the same time, Facebook’s tool for advertisers uses the cookies that it has collected from its 2,000 million users to be able to segment audiences.
Privacy VS personalization
This model will end in 2023. But on Facebook, as in the rest of the sector, they are clear that they want the wheel of personalized advertising to keep rolling. “There’s nothing wrong with non-personalized ads, it’s just that they create a completely different environment,” says Savage. “The interaction is less. It is also more difficult for businesses that advertise, especially those specializing in niche products. ” From Facebook they wield a study prepared by IAB, the influential employer association of digital advertisers, according to which 75% of the Europeans surveyed would choose the current experience of the internet over an internet without personalized ads.
The company assures that those who will really suffer the damages of the change of model are small companies, both advertisers and publishers: those who have little data and depend more on third parties.
So that everything remains the same when the cookies From third parties, new formulas are needed to amass user data in other ways. This need is joined by another issue, increasingly relevant for Internet users: respect for privacy. “The industry is evolving. We understand that people want to know who is collecting data and for what. We want technology not to force us to choose between privacy and personalization, ”says Meera Krishna, Facebook’s head of marketing, privacy and artificial intelligence for Europe.
Technology and cooperation
That can be achieved, they emphasize from Facebook, with the appropriate technology and with cooperation. The company is working on three main tools. The first is Multi-Party Secure Computing (MPC). The idea is to be able to analyze data that has never been directly accessed. If Facebook has data about a person and another company has data about the purchases that this person has made, both parties can obtain information about the missing pieces of the puzzle without having direct access to the private data that contains it.
They also investigate the so-called blind signatures, which certify that a click has been made on a banner advertising, although without identifying the author, and learning about the device itself, which anonymously adds everything that happens on the devices with which we connect to the internet.
As these strategies, still in development, sound like Chinese to non-experts, Savage provides a practical example to understand where the shots go. Let’s imagine that we want to know the average age of a certain group of people, but without knowing how old each individual is. One option would be to generate two figures for the age of each person whose sum gives the exact figure and host them on two different platforms, A and B. For example, 46,890,327 and -46,890,272 add up to 55. If we have both figures, the hosted on server A and server B, the desired data is obtained, but it is impossible to find it only from one of them. Let’s imagine that we repeat the process with all the individuals in the group, which can be thousands. If we add all the figures for each of the servers, A and B, and divide the result by the number of individuals, we obtain the average age of the group. Without anyone’s exact age being saved on any of the servers.
A system like the previous one only works if the different operators share their data. “Right now the future is not clear. But we believe that if we all come together and have a common and shared vision of what type of internet we want to have in the future and if we collaborate to make that happen, I believe that these technologies will help us to achieve it, ”said Savage.