Why a Privacy Framework Backed by Meta and Mozilla Is Gaining Traction

Leaders from the world of publishing, technology and advertising gathered inVancouver last week to craft a more privacy-safe web. The World Wide Web Consortium (W3C), a body charged with rewriting internet standards, brought the group together. One point of discussion was the future of attribution—where advertisers link the viewer of an ad to those who took a desirable action, like going to a website or making a purchase. In a world without third-party cookies and guarded by increased privacy regulation measuring the impact of advertising without leaking identity data will be tougher. Of course, Google and Apple, the tech giants behind the two most popular browsers, have submitted their visions for privacy-safe attribution protocols. But alongside these ideas, another framework driving significant discussion at the W3C Conference was Interoperable Private Attribution (IPA), developed by both Mozilla and Meta, two sources told Adweek. For IPA to gain any traction is significant given that its backers hold more of a David-sized positioning in the world of web browsers: in August 2022, Chrome had 65.5% market share and Safari a 18.8% market share compared to Mozilla Firefox’s 3.2% (and Meta doesn’t have a browser), according to StatCounter. But IPA’s lack of incumbency might be an advantage, said a publishing ad-tech expert who attended the conference—and spoke on the condition of anonymity because they lacked authorization to speak to the press. “The big problem is a trust problem,” the source said. “There is nothing wrong with Google’s [approach] and Apple’s approach, but they do depend on you trusting that Google’s or Apple’s code is effective and nonbiased, even if it’s open sourced and people don't trust it.” The source added that there are a lot of people generally leaning in IPA’s favor, noting that the industry still has a long way to go before adopting a standard, which will likely be a combination of a variety of solutions. What is IPA? IPA is a framework for attribution without tracking users across the web. A key facet of the protocol is its basis in multi-party computation, where the process of matching user and action is shared between multiple servers. In other frameworks, all computing happens in one browser. User data is broken up like pieces of a puzzle and encrypted, in something called a match key, which anyone on the web can operate. Parties only have access to data if their match keys correspond to one another—indicating the same user visited a publisher as, say, made a purchase on a website. However, once an advertiser gets access to this data, it would be batched and aggregated, meaning the user’s identity is never revealed. “No one owns the whole process,” the publishing ad-tech source said. Finding consensus A second ad-tech source, who requested anonymity to speak freely, said that sometimes when the tech companies sit down to talk about standards, gridlock quickly ensues. But the spirit of this conference, especially in the Privacy Advertising Technology Community Group where these conversations take place, was one of compromise. “There was no, ‘How are we ever going to resolve these differences?’” the source said. Indeed, there are still many open questions. For example, Google Chrome engineer Charlie Harrison inquired about IPA at the PATCG meeting, according to official meeting minutes. “Which events are acceptable inputs to the API?" Harrison said. "[Apple's Privacy Click Measurement] only allows clicks, [Android's Attribution Reporting API] allows both clicks and views.” A Meta spokesperson said that conversations at the PATCG meeting centered more on how the proposals differ than the technicalities of the proposals themselves. “We are encouraged by the progress made in the W3C meetings this week—and the common desire among industry participants to make progress toward a web standard,” said Erik Taubeneck, research scientist at Meta. The group is still far from landing on any one solution—and it’s far from certain if IPA will win or even be the grounding framework. “We believe it's too early to tell whether the IPA proposal is the right solution for the ecosystem—and we look forward to exploring it alongside other proposals from the industry,” a Google spokesperson said. And even those outside the major tech companies aren’t sold on the viability of IPA or any privacy-centric attribution framework. Don Marti, vp of ecosystem innovation at Café Media, said that the attribution framework aims to make all media comparable, but may not take into account that each media have different anti-fraud measures—making it hard to compare results. "It’s hard to say whether the ability to do attribution across media will be possible to do in a fair way," he said. Google’s Privacy Sandbox Origin Trials Are Officially in the Experimental Stage