Data collaboration isn’t new. It’s an old practice that just changed in nature and ambition.

By Michael Froment, CEO and Founder of Commanders Act

Anyone who’s worked in the digital ecosystem for more than fifteen years has already lived through several waves of what used to be called “data sharing.” The earliest forms of collaboration between players were rudimentary: CRM files exported manually and sent to a media partner as hashed email lists for exclusion or targeting, or DMP audience segments built on third-party data and shared between publishers and advertisers in a still-nascent programmatic ecosystem. Or retail data co-ops, where several non-competing retailers pooled their point-of-sale data to build richer customer profiles.

These practices worked, imperfectly, in a world where data flowed freely, where consent was a formality, and where third-party cookies let any player observe user behavior across the web.

That world no longer exists. And that’s precisely what explains the deep, structural resurgence of data collaboration.

Why data collaboration is back, in a radically different form

The death of third-party cookies, GDPR, Apple’s ATT, identifier fragmentation: it all points to the same conclusion. Data that used to flow freely between ecosystem players no longer flows. Every organization is left with its own first-party data, locked inside its own environment, unusable beyond its walls.

Paradoxically, it’s this closure that creates the need for a new form of collaboration. A brand holds valuable knowledge about its customers: purchase behavior, loyalty, value. A publisher holds a qualified, contextualized audience. A retailer holds transaction data of incomparable richness. Separately, these assets have value. Together, they enable use cases neither party could build alone — provided each accepts this logic of sharing, regardless of how much data is otherwise available.

The topic’s resurgence is also fueled by a recent, significant news event: Publicis’s $2.2 billion acquisition of LiveRamp in May 2026 is, at its core, a bet on data collaboration as the central infrastructure of tomorrow’s digital marketing. Publicis is buying the ability to organize secure data exchanges between advertisers, publishers, and platforms — at scale, within a framework that respects privacy.

What a data collaboration clean room actually is today

Modern data collaboration rests on a principle fundamentally different from its historical forms: data no longer gets “shared” — it gets cross-referenced in a neutral environment, without ever leaving either party’s zone of control.

Technically, this takes the form of “clean rooms” — secure analysis environments where two organizations can query each other’s data without either seeing the other’s raw data. The advertiser brings its customer base. The publisher brings its audience data. The clean room computes overlaps, coverage rates, and performance indicators, and returns only aggregated results — never individual-level data.

Google has its clean room, Ads Data Hub. Meta has built similar capabilities. Independent players (LiveRamp, Habu, InfoSum) offer neutral, multi-party environments that let competing players collaborate on shared use cases without exposing their proprietary data.

The use cases are concrete and numerous: measuring a campaign’s real impact by cross-referencing a publisher’s ad exposures with an advertiser’s sales data — an issue we touched on recently in our piece on Conversion APIs. Building enriched activation audiences by combining CRM data with a partner’s behavioral data. Optimizing retail media investment by cross-referencing a retailer’s sales data with an advertiser’s purchase-intent data. Reducing exposure duplication across partners by computing audience overlaps without ever sharing the underlying files.

Why a brand should adopt data collaboration

The reasons are both defensive and offensive.

Defensively, data collaboration is a response to signal fragmentation. In a world with no third-party cookies, no IDFA, and a restrictive GDPR, every player’s first-party data is necessarily partial. Nobody sees the entire customer journey. Collaboration rebuilds a more complete view without violating privacy principles.

Offensively, it unlocks targeting, measurement, and personalization capabilities unreachable solo. A brand that can cross-reference its customer base with a major publisher’s or retail partner’s data gains insights it could never build alone. It can reach audiences that resemble its best customers in premium environments, with deterministic rather than probabilistic precision. Amazon, for instance, offers highly useful geomarketing capabilities: where are products in my category actually selling? Where am I underperforming?

In a tough economic climate, where every euro of media budget has to prove its worth, data collaboration lets you allocate budget toward the audiences most likely to convert, with impact measurement grounded in real data cross-referencing rather than approximate models.

What a brand needs to watch out for

Data collaboration isn’t risk-free, and the market’s enthusiasm shouldn’t obscure some fundamental precautions.

Dependency on the partner’s environment. A proprietary clean room — Google’s, Meta’s, Amazon’s, or a major retailer’s — is an environment you don’t control. The rules of the game can change, access can be conditional, results can be filtered according to the operator’s interests. Stated neutrality doesn’t guarantee actual neutrality.

Regulatory risk. Even within a clean room, the data you bring must be legitimately collected with the appropriate consent. A technically secure data cross-reference can still be non-compliant if the legal bases for processing aren’t properly established on both sides.

Strategic risk. Data collaboration means revealing, at least partially, the structure of your customer base to a partner, even if individual-level data is never exposed. The insights you let the other party build about your audience carry competitive value. Choosing who you collaborate with, on what scope, with what contractual protections, is as much a strategic decision as a technical one.

Value asymmetry. In many collaborations, one party brings more valuable data than the other and gets less benefit from it. Clearly defining what each party contributes and receives in return, and ensuring that balance holds over time, is a condition for the collaboration’s durability.

What data collaboration requires from your infrastructure

At Commanders Act, we see that the brands best positioned to benefit from data collaboration are the ones that first built quality first-party data — cleanly collected, consented, controlled, enriched, governed. That’s our day-to-day focus for our clients: you can’t collaborate well with data you don’t own well.

A mature server-side infrastructure is the natural prerequisite for data collaboration: it guarantees that the data you bring into a clean room is complete, reliable, consented, and enriched. Poor, incomplete, or badly governed data doesn’t produce better insights just because it’s cross-referenced with someone else’s poor data.

Data collaboration is a real promise. But like every promise in digital marketing, it depends first on the quality of what you bring to it.

About the author

Go further