S2E16: Leading with Trust, with ID5's Mathieu Roche

Full analysis

Mathieu Roche, co-founder and CEO of identity infrastructure company ID5, went on Geoff Wolinetz and Greg MacDonald's leadership podcast and made a pitch that should worry every operator who leaned on one vendor for identity. His argument: Publicis buying LiveRamp hands the whole market a reason to shop for an alternative. He's talking his own book, and hard. But the underlying move he's describing is real, and it's the thing to weigh.

What's actually being decided here is whether operators who parked their identity resolution with LiveRamp now need to diversify. That decision is easy to undo at the pilot stage and hard to undo once you've rewired match keys, consent plumbing, and reporting around a new provider. Nothing sets a hard deadline. The pressure is competitive, not regulatory.

One caveat up front. The summary flags that a completed Publicis acquisition of LiveRamp is unverified in established reporting. Roche is the single most conflicted person to be narrating it. Treat the deal as his claim, not as fact, and everything below as conditional on it being real.

The Market Analyst

If Publicis owns LiveRamp, the pitch writes itself. LiveRamp sits in the middle of the industry as neutral connective tissue: the place brands, publishers, and platforms match their data without handing it to a competitor. Put a holding company on top and the neutrality claim gets awkward. A rival agency does not want its client data flowing through Publicis-owned pipes. That is the opening Roche is selling. In plain terms: the plumbing everyone trusted because nobody owned it now has an owner, and the owner competes with half its customers. The question is whether that discomfort converts to switched contracts or just to grumbling in renewal meetings.

The Skeptic

Roche has an obvious incentive to make this sound bigger than it is. "Oxygen in the room" is not a signed contract. Two things have to be true for his story to pay off. First, the deal has to be real and closed. Second, discomfort has to beat switching cost, and switching identity infrastructure is not a weekend job. LiveRamp is embedded in match keys, clean-room connections, and years of onboarding. The sub-50% cookie figure is his too, and its definition is fuzzy. Directionally fine, precisely unverifiable. A CEO whose whole business depends on cookies dying tells you cookies are dead. Note the source.

The Operator

Say you buy the argument and start diversifying Tuesday morning. What breaks first? Your reporting. Run two identity graphs and your match rates, your audience counts, and your attribution stop reconciling, because each vendor recognizes a different slice of users. Roche's own framing helps here: the post-Sandbox world is a patchwork across Safari, Windows browsers, apps, and CTV. Adding a second identity vendor does not clean that up, it adds another set of numbers your team has to explain to clients. The 90-day surprise is not the integration. It's the quarter your dashboards disagree and nobody can say which one is right.

The Customer / End User

Roche's consent model is the part regulators will poke at. His chain of trust runs publisher to ID5, not user to ID5. The user trusts CNN or Spotify, the publisher trusts ID5, and Roche says that delegation is legitimate as long as media owners behave. Maybe. That is exactly the arrangement US state privacy enforcers and EU regulators are pressure-testing right now. If the model breaks, it breaks for every identity vendor riding on publisher consent, not just ID5. Roche says ID5 is happy to meet the higher European bar even where the market doesn't demand it. Fine. Operators relying on publisher consent to cover downstream vendors carry that exposure, and that is most of them.

The CFO

The real cost of diversifying is not the second vendor's fee. It's the integration engineering, the parallel-running period where two systems report different numbers, and the client conversations explaining the gap. Against that, the benefit is optionality: not being captive to a provider your competitor now owns. That is worth something, but it is insurance, not revenue. Most operators will do what CFOs always do with insurance. They'll run a small pilot, keep LiveRamp live, and delay the real switch until a renewal or a price change forces the math. Roche gets a test budget. He does not get the whole account.

Where the council splits

Two real disagreements. The Market Analyst sees a structural opening, because a holding company owning neutral plumbing genuinely changes the trust calculus. The CFO and Skeptic see switching cost and conflicted testimony swallowing most of that opening, so the opening produces pilots, not migrations. Second split: the Customer lens says the consent model Roche leans on is the shaky part of the whole independent-identity story, while everyone else treats identity resolution as a plumbing question. If regulators go after publisher-delegated consent, the plumbing debate is beside the point.

What it hinges on

Three things. Is the Publicis-LiveRamp deal actually closed on the terms Roche describes. Does buy-side discomfort with holdco-owned infrastructure beat the cost of rewiring identity. And does publisher-delegated consent survive the next round of privacy enforcement. The council leans skeptical on the near-term commercial impact and takes the fragmentation point seriously. Verify the deal before you touch your roadmap. If it's real, run a scoped pilot with a second identity provider, but keep your incumbent live and reconcile the reporting before you commit anything hard to undo.

This is a moderate-impact episode. The framing is useful. There are no new facts.

Prediction: At least one of the independent identity vendors (ID5, The Trade Desk's UID2, or LiveRamp itself) will publicly announce a new publisher or platform partnership explicitly framed around neutral, holdco-independent identity by the IAB Annual Leadership Meeting in late January 2027.

Confidence: Medium. The competitive opening is real, but timing depends on deal close.

Why: If a holding company like Publicis owns LiveRamp, rival agencies and independent publishers have a concrete reason to avoid routing their data through a competitor's pipes, and independent identity vendors have every incentive to market against exactly that discomfort. Roche is already running this pitch publicly, which means the messaging is live even before contracts move. The opposite outcome, no vendor making a neutrality-framed announcement, would require the entire independent identity category to sit on the single best marketing hook it has been handed in years, which vendors competing for the same displaced accounts will not do. The weaker link is timing and whether the LiveRamp deal is fully closed, not whether the pitch gets made.

Revisit by 2027-02-01: We're right if an independent identity vendor announces a partnership or product positioned on holdco-independent, neutral identity by the IAB Annual Leadership Meeting in late January 2027. We're wrong if no such neutrality-framed announcement surfaces by then.