AdTech in 2022: Uncertainty, Hope, and Painful Trade-Offs
Earlier this month I wrote a summary of the ideas laid out during ExchangeWire’s ATS Madrid, together with my own conclusions. I am re-publishing it here in a slightly shorter form.
Let’s start with the one sentence that I found most descriptive of a gathering in which publishers, ad tech vendors, media agencies, and advertisers (these last two in much smaller proportion) debate the future of media, addressability, relevant ads, and the open market in a privacy-first world:
Nobody knows what they’re doing
These words were uttered by an experienced member of the digital advertising ecosystem. But, rather than invalidating the entire effort (an industry coming together to talk, share, and ponder new ideas), they provide a great introduction to what I believe is the scenario at play:
We have the publishers on one side. They feel the time has come to take control of their products/experiences, their audiences, and their fortunes. Beyond the uncertainties of a subscription-based model, they are putting together a few innovative ideas around first-party data and onsite conversions (e-commerce), but seem ready to put them on hold if they can please agencies and brands for the sake of short-term revenue.
On the other side we have media agencies, as trusted partners in charge of advertising budgets. It is hard to demonstrate real value to a client when you follow the easiest path to digestible metrics and plausible results. After all, the giant walled gardens are naturally built for disintermediation — they are also few and easy to grasp (curatorship and brokerage require abundance and complexity).
Agencies are well aware of the upcoming disruption to an open market that they have navigated rather successfully, but they appear to be waiting for publishers to come up with a Plan B and, to make it even harder, they are expecting such alternatives to preserve their current planning, activation, and measurement playbooks.
This context paves the way for the lowest hanging fruit: an identity solution that works in a (third-party) cookieless browser. Shared IDs are not really meant to solve a privacy problem (in fact, hashing actual email addresses is clearly worse for privacy than using random, ephemeral third-party cookies) — but rather a technical and competitive challenge, posed by the elephants (not) in the room and stretching across both the open web and mobile app ecosystems. These have been proven to work, but are surrounded by even more uncertainty in terms of legal bases and platform support, besides the fact that they replace a common standard with a private, for-profit solution.
But, whether or not there is a tangible, growing demand for individual privacy, the legal cover for these challenges is clear and present. Furthermore, legislators and supervisory authorities are pushing the open market into a corner by:
- Closing in on valid consent, with demands for a “Reject All” button on the CMP’s first layer — ExchangeWire’s Co-Founder Ciaran O’Kane pointed out the catastrophic impact that this is already having in France or Italy on the size of the resulting samples. True, some brands expect to rely on a superior user experience to increase consent rates, but there is a thin line separating UX optimization and the use of dark patterns (which both the Digital Services Act in the EU and the lawsuits put forward by a group of Attorney Generals in the US are hoping to do away with).
- Making sure that no data is sent to the US, after the Belgian DPA’s resolution -invalidating the IAB’s Transparency and Consent Framework- woke the entire Ad Tech ecosystem to the post-Schrems II world
- Forbidding RTB when children and special categories of data are involved (as per the EU’s Digital Services Act, and following self-regulatory measures on the part of the largest walled gardens)
- Providing an opt-out of personalized advertising (as reflected in the EU’s Digital Services Act and some State-level privacy laws in the US) or suggesting an all-out ban of RTB, both in the EU (through the EDPS’ commentary to the Digital Services Act) and the US (through the proposed Banning Surveillance Advertising Act).
And while agencies seem oblivious to these developments, as if taking some sort of miracle for granted, publishers (who bear the burden of declaring a legal basis on behalf of everyone else) are feeling the heat. But they are also waking up to the fact that the story could have a different ending this time around, and that the world in which their brands, premium content, and quality audiences dissipate in an ocean of commoditised inventory may have come to an end.
It seems like that would not be the case if Google Chrome’s Privacy Sandbox becomes the new status quo, as all sorts of websites, regardless of their quality or veracity, will have access to the same Topics API on people’s browsers. Definitely better for privacy than the former FLoCs proposal, and surely an improvement over third-party cookies and, needless to say, unified ID solutions. But still suboptimal with regards to preserving that differentiation on the supply side.
Of course, the question remains open as to whether such premium content does indeed deserve premium prices. As discussed on one of the panels, this is the publisher’s internal business case for decreasing available inventory in the process of gathering first-party data and piling up subscribers. And the constraints on cross-site measurement make it very hard to back up with solid numbers.
So I could not help thinking of Twitter, and its systemic inability to catch up to Instagram, TikTok, or Snapchat in terms of advertising revenue. It would seem like publishers face a very similar dilemma: an audience that is not looking for passive entertainment (wasting time!), perfectly suited for continuous interruptions. They are instead expecting to actively engage with news, analysis and opinion pieces — no matter how premium that inventory, it becomes the least suitable place for an interruption, relevant or not, contextual or behavioral, subtle or intrusive.
In other words: Is it possible that the junk inventory that ad fraud researchers have been denouncing all along did in fact deserve to be pooled together with that of a Condé Nast or Prisa?
Which leads to something that ExchangeWire’s CEO Rachel Smith kicked off the event with: “everyone is a walled garden” — with first-party data solutions reigning supreme. How would publisher-specific environments work? How about retail media? And these questions were also very present.
The representatives of Spain’s two largest publishers explained that their newly found ability to better understand their audiences (in order to turn readers into subscribers) is now providing them with a powerful source of insights and segmentation vis a vis the advertising ecosystem. Third-party solutions are also making this latter effort easier, and I was left wondering about the potential disadvantages of tackling these two endeavors separately.
Going further into my own personal analysis, I reckon that quality content should fly clean and separate from demand generation. I just don’t think they can coexist in space and time — as they naturally did in the days of printed media and linear TV. On their part, retail media players deal with demand discovery instead (just as search does), and all incentives seem perfectly aligned on their end.
All in all very dense, requiring difficult choices at all levels and, as expected, exposing the short-term/long-term trade-off that most organizations are really terrible at handling.
In the meantime, it definitely seems like those who were not there (Google, Apple, Meta, data protection regulators, competition authorities…consumers!) are the ones that will decide AdTech’s future.