Update: Public ad-tech exodus accelerates: IAS, LiveRamp, Criteo eyed for private deals

Full analysis

What's new since we last covered this: LiveRamp deal now closed/Publicis acquisition confirmed; IAS and Criteo PE exits are new developments.

Three big ad-tech names are heading for the exits at once. IAS went private with Novacap for just under $2 billion. LiveRamp, the data-onboarding company that ties offline customer files to digital ads, agreed to sell to Publicis with a close expected by end of 2026. And Criteo, which just posted a 14% revenue drop to $428 million for the quarter, is now the reported target of PE firms Vista Equity Partners and Quinti Capital. The question for operators: is this a structural thinning of the independent middle layer, or PE picking over cheap distressed assets?

Reversibility: Type 1 for the companies involved. Once LiveRamp is a Publicis asset or Criteo is a Vista portfolio company, the neutral-utility version of those businesses is gone for good. For the operators watching, the decision is Type 2: identity diversification and vendor hedging can be adjusted quickly. Forcing function: the LiveRamp close by year-end 2026, and Criteo's board is clearly in play now.

The Market Analyst. The obvious trade already happened. Standalone mid-tier ad-tech has been discounted for 18 months, so IAS at under $2 billion and a wounded Criteo aren't surprises. The repricing worth watching is upward, on scarcity. If LiveRamp becomes a Publicis-owned asset, every buyer that isn't a Publicis client now needs a neutral identity source, and that hands ID5 and InfoSum a scarcity premium they didn't earn on merit. The Trade Desk wins quietly too: fewer credible independent data partners means less friction for its Unified ID 2.0. In plain terms, when the neutral middlemen get bought by one team's owner, the remaining neutral players become more valuable.

The Skeptic. Three deals are a press cycle, not a wave. IAS was verification commodity. LiveRamp's holdco logic was obvious years ago. Criteo has been a value trap since Apple's tracking changes gutted retargeting. The three companies share one thing: weakness. PE is bottom-fishing at depressed prices. For an actual exodus you'd need healthy, growing firms exiting, and you don't have one. The Trade Desk and Magnite at scale aren't going anywhere. This is triage. In plain terms: the weak hands are folding, and folding weak hands is what markets are supposed to do.

The Operator. The people cuts start before the ink dries. Finance and investor-relations headcount at mid-tier independents is Q3/Q4 roadkill. The deeper problem is lock-in. LiveRamp's onboarding pipelines will serve Publicis client priorities first once Publicis owns them, which means latency and reprioritization for every non-Publicis buyer that depends on RampIDs. If Criteo goes to Vista, expect a roadmap freeze while the cost-compression playbook runs. Retargeting SLAs slip quietly before anyone files a ticket. Procurement at non-holdco agencies should be adding a second identity source now, not after the 12-month close. In plain terms: your vendor's new owner has other customers ahead of you.

The Customer / End User. Put yourself in the seat of an independent agency or a mid-size publisher. Your identity spine just got adopted by a competitor's parent, and your verification vendor now answers to a PE clock. Nobody asked for this. The buyer who valued LiveRamp precisely because it was Switzerland now has to assume it won't stay neutral. That doesn't mean rip and replace tomorrow, but it does mean you stop treating any single onboarding partner as permanent infrastructure. In plain terms: the neutral plumbing you built on is quietly changing owners, and the new owner has a favorite.

The Strategist. This is compression, not a cycle. The mid-tier public layer, verification, onboarding, retargeting, is being absorbed into holdco stacks or PE harvesting vehicles because none of it had a moat that justified a standalone public premium against Google, Meta, and Amazon's data gravity. In two to three years the independent open-web stack thins to a handful of scaled SSPs and DSPs, and everything else is a feature inside a holdco or a zombie. The casualty is neutral identity infrastructure as a shared utility. In plain terms: the parts of the open web that were owned by nobody are becoming owned by somebody.

Where they part ways

The Skeptic and the Strategist are having the real fight. Skeptic says three deals are just weak companies clearing out, and the strong independents stand untouched. Strategist says the weakness isn't company-specific, it's the whole mid-tier layer losing its reason to exist as standalone businesses. That's not a mood difference. It's a genuine disagreement about whether the middle of the open web has a durable moat or was always a feature waiting to be absorbed.

The second split is on the LiveRamp-Publicis close. The Market Analyst treats it as a gift to whoever stays neutral. The Operator treats it as a slow-motion problem for every non-Publicis buyer wired into RampIDs. Both can be right: neutral rivals get more valuable precisely because the incumbent is getting less neutral.

What it hinges on

Two things. First, whether neutrality is a real, priced asset or just a nice story. If buyers actually pay a premium to route identity through someone with no agency parent, then ID5, InfoSum, and the independent DSPs gain. If buyers shrug and keep using RampIDs because switching is painful, the neutrality premium is a slide, not a business.

Second, whether Criteo actually sells. That's the tell that separates Skeptic from Strategist. A cheap verification buyout and an obvious holdco tuck-in don't prove a structural thesis. A distressed retargeter getting taken by Vista at a beaten-down price would confirm PE sees a whole category worth harvesting, not one asset worth fixing.

The council leans toward the Strategist on direction and the Skeptic on pace. The middle layer really is thinning. But it thins one folding hand at a time, not in a stampede. What operators should de-risk now is single-vendor identity dependence, because that exposure changes owners faster than contracts renew.

The call

Prediction: Criteo will announce a take-private agreement with a private-equity buyer (Vista or another PE firm) before its Q2 2027 earnings report.

Confidence: Medium. Named PE suitors, a 14% revenue slide, and a board already in play.

Why: Criteo just posted a 14% annual revenue decline to $428 million, and two PE firms, Vista and Quinti, are already reported to be circling, which is the setup that precedes most take-privates rather than a passing rumor. The retargeting core has been shrinking since Apple's tracking changes, so there's no organic growth story that argues for staying public and fighting volatile markets. When a shrinking business with a broken moat has named buyers and depressed shares, the board's cheapest exit is a private deal, and IAS going to Novacap this year shows PE is actively writing checks in exactly this range. The opposite outcome, Criteo staying independent and public, requires a growth reversal there's no sign of and a management team choosing to keep grinding in the market that punished them.

Revisit by 2027-05-15: We're right if Criteo signs a definitive take-private deal with a PE buyer. We're wrong if Criteo is still an independent public company with no signed agreement by its Q2 2027 print.