Online advertising is worth $300bn (€251bn) annually. And it’s currently being torn apart and rebuilt behind the scenes.
The basis of the ads that track you around different websites is the third-party cookie — a small file that tells advertisers, social media platforms and others where you’ve been. But the cookie’s days are numbered. They have been for some time. The demise of the cookie was recently hastened when Google announced that it would stop using any tracking technologies that uniquely identify web users as they browse the internet.
This is a big deal from the company that has arguably benefited the most from the cookie — Google accounts for around half of all digital advertising — and will have direct ramifications for others. So what comes next and how will it affect digital marketing overall and the businesses that rely on online ads?
Google’s focus is now on what it’s calling a “privacy sandbox” — methods of targeting ads without collecting information about individuals. Google and others are proposing a host of bird-themed acronyms such as PIGIN (Private Interest Groups, Including Noise), Swan (Storage With Access Negotiation) and Pelican (Private Learning and Inference for Causal Attribution). What is it with big tech and ridiculous acronyms?
Anyway, the most prominent of these new services is called FLoC — or Federated Learning of Cohorts. Google’s plan with FLoC is to anonymise and aggregate your recent browsing activity into a behavioural label that will be shared with websites and advertisers. It’s less of a cookie and more like a cookie jar, into which advertisers will reach with their eyes closed. Scale matters for online advertising, so Google is proposing that a central administrator will prevent any cohort from being too small. Small cohorts will be combined with other, similar cohorts until enough users are represented in each one.
FLoC cohorts will be recalculated on a weekly basis, each time using your data from the previous week’s browsing, making FLoC cohorts less useful as long-term identifiers. Plus the assignation of cohorts will be done on your browser, which means your information is kept locally, rather than stored on a remote server. Sounds good, right? So good, in fact, that Google is telling advertisers that FLoC is 95pc as effective as the third-party cookie based on recent trails. But there are a host of concerns.
There are privacy concerns. The Irish Council for Civil Liberties’ Dr Johnny Ryan wrote: “Google has not yet provided sufficient information for one to judge whether its new advertising system will end the enormous data free-for-all among thousands of companies active online advertising industry.” The Electronic Frontier Foundation has called FLoC the “the opposite of privacy-preserving technology". Plus it has pointed out that FLoC groupings may unwittingly create cohorts of vulnerable users based on visits to websites related to substance abuse, financial hardship, or support for survivors of trauma.
There are concerns about Google’s market dominance. Marketers for an Open Web said: "this is a monopolistic player attempting to consolidate its dominance by degrading the Open Web using privacy and collaboration as a veil of legitimacy”.
And there are concerns for the marketing sector and publishers. The Interactive Advertising Bureau (IAB) has warned that the online advertising industry isn’t prepared for the loss of third-party cookies and publishers risk losing digital advertising revenue to walled gardens, as the open web becomes a more confusing place to target potential customers. FLoCs will only work on Google’s Chrome browser, making things even more complicated.
But here’s the real problem with FLoC and the other avian acronyms; they aim to replace existing systems of advertising not improve them. It’s an area that could do with improvement. Last year, a study from the University of Baltimore found that global online ad fraud cost advertisers $35bn (€29bn) in 2020 .Another study from advertiser body ISBA and PwC in the UK found that 15pc of advertiser spend — around one third of supply chain costs — was completely unattributable. The same study also found that the sites that serve ads only receive half of the cost advertising — the other 50pc of the advertising costs evaporate in what P&G’s Marc Pritchard memorably called a “murky media value chain”.
Some brands are even questioning the value of their ad spend. Airbnb, for example, cut its marketing spend by 58pc or $662m in 2020 due to the pandemic. Did traffic plummet? Hell no. The company attracted 95pc of the traffic that it did a year earlier when it spent $1.14bn.
And while marketers should always question their ad spend, publishers should also be wary of signing up to FLoC technology. We all know that advertising is part science and part art. However, digital advertising and its reliance on the click has deprioritised the art in favour of the science. And Google’s approach to replacing cookies continues this imbalance. This is all very well in certain domains such as search and social advertising, where targeting based on intent to purchase or known interests mean that science and analytics belong in the fore. But this doesn’t necessary hold true on the likes of publisher sites, which should be making cultural relevance their USP and focusing on first party data strategies.
But let’s remember, the impetus for Google’s introduction of non-tracking technologies isn’t to fix transparency, address fraud or help publishers. It is to minimise risk around legislative changes and market dynamics. New laws in Europe and California are raising awareness of privacy and allowing more of us to control what data we share and how it can be used. Google is well within its rights to recreate its cookie-based cash cow for a more privacy-centric web. The problem is, given Google’s prominence, most marketers and publishers will have no real option other than join the FLoC.