Facebook’s privacy calculus

Yesterday, Facebook published a blog post detailing the work it is doing to transition its advertising infrastructure away from “individual third-party data.” In the blog post, Facebook cites platform policy changes and an evolving regulatory landscape as instigating this change — the post also champions the utility of privacy-preserving technologies in delivering effective, performant advertising in a way that protects consumer data from abuse. Two of the technologies that are specifically cited in the post are on-device data processing, which I discussed in a recent Twitter space, and secure multi-party computation. A Facebook advertising executive’s interview with The Verge, published alongside the blog post, reveals Facebook’s ambitions with its privacy transformation in more detail.

Since Apple’s App Tracking Transparency (ATT) policy was first introduced last June, a number of interesting and technologically impressive mechanisms that safeguard consumer data have been proposed from all corners of the advertising ecosystem. Most of these solutions rely on some combination of differential privacy, on-device data processing, or cryptography. And inherent in the existence of these solutions is a recognition of the tradeoff between privacy and utility with digital products: that there exists an optimal point on the privacy-utility spectrum that can be achieved through technological sophistication.

Facebook’s blog post seems to likewise acknowledge as much: tools can be built that provide advertising efficiency using only first-party data and which satisfy the restrictions of ATT (and likely any other conceivable privacy restrictions to come). Facebook’s goals, as stated in the blog post, are noble, and the technology it discusses seems eminently viable. But as I state in the above Twitter thread, I believe that Facebook’s pain related to ATT is only going to be felt in the short term, assuming it can execute against its medium-term ambition of creating a content fortress, meaning all user conversions take place within its content ecosystem and are thus fair game for use in ads targeting, as with the recently-launched ad product in the Instagram Shop tab. And of course, ATT is probably irrelevant to Facebook’s longer-term ambition of becoming a metaverse company.

But what if the platforms and regulators that control the destiny of digital advertising don’t recognize a privacy-utility tradeoff — what if their approach to moderating user privacy resembles something more akin to religious zeal than an economically-oriented consideration of consumer welfare? Or what if, cynically, privacy is being wielded as a weapon by platforms to preserve control of content distribution or to capture share of the advertising market? In either (or both) of these cases, the technology deployed against the compound goal of preserving advertising efficiency while safeguarding user data doesn’t matter: privacy policies can be polished and honed and made more precise to prohibit any approach that isn’t deemed aligned with whomever’s interests are being protected. In other words, the goal posts can move.

Facebook’s privacy calculus seems to be that, regardless of how aggressive platforms or regulators become in pursuit of perfect privacy, advertising technologies only need to withstand the pressures of the next few years. Like a billionaire who built a climate bunker in New Zealand as a temporary stop on the path to going interplanetary, so is implementing tooling that is consistent with any recognition of a privacy distinction between first- and third-party conversion data. Once the walled gardens become fortresses in the medium term (and potentially metaverse companies on a much longer timeline), then the substance of the privacy discourse becomes largely irrelevant to the operators of the largest ads ecosystems.

Photo by Samuel Ramos on Unsplash