DoubleVerify, LiveRamp, IAS taken private; Criteo exploring deal

Full analysis

Three verification and identity companies left the public markets this year. DoubleVerify and Integral Ad Science, both of which check whether ads actually ran next to safe content and reached real people, went private. So did LiveRamp, which connects customer data between companies and stitches identities together. Criteo, the retargeting company that follows you around the web with ads for the sneakers you didn't buy, is now rumored to be looking at the same exit. The bankers say acquirers see bargains, since these names trade at a fraction of what they fetched five years ago.

What's actually being decided here isn't any single deal. It's whether the standalone public ad-tech company is still a viable business model, or whether the whole middle of this industry gets absorbed into bigger data and software clouds. That's hard to undo once it happens. When a verification tool becomes a checkbox inside Salesforce, it doesn't come back out as an independent vendor. No hard deadline, but Criteo's process and the next round of ad-tech earnings set the clock.

The Market Analyst. The public ad-tech multiple got crushed starting with Apple's privacy change and never recovered through the cookie mess. These take-privates confirm the floor. The interesting consequence is what it does to the survivors. With the muddled middle gone, the pure-play independents left standing, The Trade Desk, AppLovin, Magnite, PubMatic, get a clarity premium. Buyers of these stocks no longer have to guess who gets picked off next at a distressed price. That overhang clears. In plain terms: when the confusing companies leave the stage, the ones remaining look more valuable simply because you can finally understand what you own.

The Skeptic. Four exits looks like a wave. It's really four companies that missed their growth targets getting bought cheap. For this to be strategic instead of financial, the acquirers have to bolt complicated, relationship-heavy ad-tech onto enterprise software without wrecking it. That's a graveyard. Salesforce bought Krux in 2016 and the data platform is now a footnote. Adobe bought TubeMogul and the buying tool vanished. And the "OpenAI is circling" line is banker fan fiction. OpenAI does not need a retargeting engine. In plain terms: cheap is not the same as smart, and most of these integrations quietly die.

The Strategist. Look at who's buying, and the story changes. Salesforce, Adobe, Databricks, Snowflake. They aren't buying ad-tech. They're buying identity graphs, measurement rails, and clean-room plumbing to fight AWS and Google for the enterprise data stack. Verification and identity stop being products you buy from a vendor and become features inside a CRM or a data cloud. In plain terms: the thing you used to purchase as its own tool becomes a line item in software you already pay for. That makes The Trade Desk and AppLovin's independence rarer, and rarer usually means more valuable.

The Operator. The pain lands on your desk in 60 days, not in some 24-month thesis. DoubleVerify and IAS customers now renegotiate contracts under owners with different margin math. Expect your account team to turn over once retention packages get sorted, so the person who knew your setup is gone. LiveRamp's pipes are wired into dozens of clean-room and identity workflows. If a new owner trims the product to hit numbers, those connections break, and your data plumbing breaks with them. Dual-source your verification now. Sitting on a single acquired vendor with a new board is a single point of failure you chose.

The Customer / End User. The brand-safety and measurement buyer never asked for this. They picked DoubleVerify or IAS partly because they were independent referees, not owned by anyone they were grading. Fold verification into Salesforce or a DSP and that independence gets murky. Who checks the checker when the checker is now a feature of the thing being checked? Agencies will feel this first at renewal, when the pitch quietly shifts from "best-in-class standalone" to "already bundled, why pay separately." Cheaper, maybe. Independent, less so.

Where the council splits. The Strategist says these buyers want the assets and will integrate them into something bigger and durable. The Skeptic says the same buyers have a long record of buying ad-tech and quietly killing it, and points to Krux and TubeMogul by name. Both can't be right about the same deals.

The second split is on the survivors. The Market Analyst says the independents get more valuable because the landscape simplifies. The Skeptic would note that "less competition on the public market" and "better business" are not the same thing, and a clarity premium is still a story investors tell themselves.

The third, quieter tension: the Operator and Customer both see disruption and lost independence where the Analyst sees a cleaner board. Same event, opposite feeling, depending on whether you own the stock or run the account.

What this hinges on: are these acquirers buying capability they'll invest in, or distressed assets they'll milk and let decay? The buyer list is the evidence. If it's Salesforce and Adobe, history says the products get absorbed and thinned. If it's Databricks and Snowflake, who are building genuine data platforms and don't already have a dead DMP in the closet, the integration case is stronger because the plumbing actually fits what they sell.

The council leans one way on the practical question. Whatever the strategic logic, the near-term reality for anyone running a verification or identity contract is disruption: account churn, product rationalization, renegotiation. Dual-sourcing verification isn't paranoia right now, it's basic hygiene.

On the bigger question, treat the "AI players circling" framing as noise. OpenAI does not need Criteo. The real buyers are the enterprise data clouds, and their track record with ad-tech is worse than the deal announcements suggest.

Prediction: Criteo will announce a take-private or strategic sale agreement by its Q2 2027 earnings report (roughly August 2027), and the buyer will be an enterprise data, cloud, or private-equity acquirer, not an AI lab like OpenAI.

Confidence: Medium — the process is already rumored and the distressed valuation invites a bid, but timing and buyer identity can swing.

Why: Criteo trades at a fraction of its five-year high and is now openly rumored to be exploring a deal, the same setup that just took DoubleVerify, LiveRamp, and IAS off the public market this year. Retargeting is a mature, cash-generating business with a shrinking independent rationale as cookies fade, which is exactly the profile private equity and enterprise data buyers pay distressed prices for. The AI-lab buyer story is banker speculation with no strategic logic: a model company gains nothing from a retargeting engine, so the plausible acquirer sits in the data-cloud or PE camp. The opposite outcome, Criteo staying independent and public, requires it to resist both a cheap valuation and an active rumor mill, which is the harder path given how the rest of the sector just went.

Revisit by 2027-08-31: We're right if Criteo signs a take-private or strategic acquisition with a PE, enterprise software, or data-cloud buyer. We're wrong if Criteo remains an independent public company with no deal signed, or if the acquirer is an AI lab.