For four years, the advertising industry talked about the cookieless future like a scheduled apocalypse. There were conference tracks dedicated to it. There were whitepapers. There were panels at Cannes with titles like “Navigating the Privacy-First Landscape” where people in expensive shoes discussed first-party data strategies with the urgency of people who had already figured it out and the vagueness of people who had absolutely not.
The future they were preparing for is now largely the present. And the plan — the comprehensive, industry-wide plan that four years of earnest conversation was supposed to produce — is somewhere between “nascent” and “fictional.” What we have instead is a collection of workarounds, a great deal of confidence in concepts that haven’t been tested at scale, and an attention economy that is doing what it has always done: charging forward without knowing who it’s charging toward.
What We Were Promised and What We Got
The narrative was straightforward: third-party cookies die, privacy becomes the default, and the industry responds by building better, more consensual, more meaningful relationships with consumers. First-party data would replace surveillance. Contextual advertising would make a comeback. Clean rooms would clean everything up. The consumer would be better served. The advertiser would be better informed. Everybody wins.
Here is what actually happened. First-party data is real and valuable and also extremely hard to build, particularly for brands that do not have a strong reason for consumers to hand over their information voluntarily. The brands that are doing this well are the ones that already had strong products, strong loyalty programs, and strong reasons for consumers to engage directly. For everyone else, “building a first-party data strategy” has meant collecting email addresses via pop-ups with aggressive opt-in copy and calling it a database.
Contextual advertising is back, yes. It is also still contextual advertising, which means that the ceiling of sophistication is roughly 2009. Showing running shoe ads next to marathon training content is not revolutionary. It is the original premise of media buying, rediscovered. The tech stack is more sophisticated. The insight is not.
The Clean Room Problem: Technical Solution, Business Question
Data clean rooms are genuinely interesting technology. The principle is elegant: two parties can compare and match their data without directly sharing it, enabling targeting and measurement without exposing raw data to either side. In theory, this solves the privacy problem while preserving the targeting capability.
In practice, clean rooms require both parties to have substantial, high-quality data; technical infrastructure to participate in the matching; a legal framework that satisfies multiple privacy regimes simultaneously; and a business relationship that makes the collaboration worthwhile. Most brands have one of these things. The largest retailers and the biggest platforms have all of them. Everybody else is watching a partnership model develop between entities that don’t need them.
The clean room conversation is, at its core, a conversation about the consolidation of advertising power into fewer, larger ecosystems. Amazon, Google, Meta — the walled gardens did not shrink in the cookieless era. They grew more valuable. Their first-party data became the thing everyone else was trying to approximate. The attention economy didn’t democratize. It concentrated.
Measurement: The Thing Nobody Has Actually Solved
Attribution was already broken before cookies became complicated. The multi-touch models were confusing correlation with causation in ways that made everyone feel better about their spend without necessarily understanding it. Last-click attribution was a known fraud that the industry continued to use because it gave clean numbers and clean numbers make good slides.
What has replaced these broken models is largely a collection of probabilistic approaches — media mix modeling, incrementality testing, brand lift studies — that are more honest about what they don’t know but significantly harder to operationalize at the pace marketing departments move. MMM requires months of data and specialized analytical capability. Incrementality testing requires experimental discipline. Brand lift studies measure something real but measure it slowly.
The result is that marketing measurement, in the cookieless era, has split into two camps: the sophisticated brands that have invested in econometric modeling and are operating with a genuinely more honest picture of their marketing effectiveness, and everyone else, who is largely optimizing toward platform-reported metrics that have the epistemological status of a grade given by the teacher whose class you’re in. The ego KPIs didn’t disappear. They just got new names.
What the Conferences Got Wrong
The cookieless future was discussed, for years, primarily as a technical problem. The technical problems are real but they were always secondary to the business problem, which is this: advertising works, to the degree that it works, because of relevance. Relevance requires knowing something about the person you’re talking to. The cookie era produced relevance at scale — imperfectly, with significant privacy costs, and with a measurement apparatus that frequently measured the wrong things — but it produced it.
The replacement infrastructure produces relevance at reduced scale, at higher cost, with greater complexity, and for brands that have the resources to invest in building it. The industry’s answer to the death of the third-party cookie has been, broadly, to make targeting and measurement more expensive and more proprietary. This is not what the conferences promised. The conferences promised a better relationship between brands and consumers. What materialized was a better relationship between brands and platforms, on the platform’s terms.
None of this was said at the panels. The panels featured case studies from brands that had done it well, which is the conference industry’s mechanism for suggesting that anyone can do it well if they simply follow the example. The storytelling was excellent. The strategy was someone else’s.
Where We Actually Are
The advertising ecosystem is messier, more fragmented, and less measurable than it was five years ago. It is also, arguably, more honest — the surveillance infrastructure that powered a decade of “personalization” was always more fragile and more ethically questionable than the industry acknowledged. The cookieless reckoning has forced a genuine conversation about what advertising is actually for and what it can actually know.
The brands winning in this environment are not the ones that found a technical solution to replicate what cookies did. They are the ones that decided to be interesting enough, useful enough, and trusted enough that consumers choose to engage with them directly. This is not a data strategy. It is a brand strategy. It requires patience, product, and a willingness to measure success over longer time horizons than a quarterly campaign report allows.
It also requires being comfortable with not knowing exactly where every conversion came from. Which is, honestly, where we always were. The cookies just gave us the comfortable illusion of precision.
If you’re building a brand in an era when nobody knows who they’re talking to, the only durable answer is to be worth talking to. That’s the whole brief. If you need something to remind you of that while everyone else debates tech stacks, the NoBriefs shop has the Spreadsheet Sloth for the meetings where the numbers are confident and the strategy is not. Wear it with the quiet dignity of someone who has read the MMM report and understood it.
Also worth reading: The Brief of the Future: Will It Disappear with Generative AI? and Data-Driven Creativity: The Oxymoron Running the Industry.
