Market Update: The AI Companies Coming for Visibility and Events
Two venture rounds worth noting this week, both at the intersection of AI, marketing, and commerce.
GeoSurge, a London-based AI company, raised a $12M seed led by Albion VC, with Play Ventures and Octopus among others. It’s data infrastructure for brand visibility inside generative AI, the AIO/GEO space that’s emerging as the successor to SEO. Not the first player here, though. In commerce, ReFiBuy raised $13M for a similar thesis. The point is that an entire new sector of AI companies is forming around one question: how do you make sure your brand is visible inside ChatGPT or Claude when someone asks for answers?
Vendelux, a New York-based firm, raised a $50M Series B from Tribeca, with HubSpot Ventures and FirstMark participating. This is AI-powered B2B event marketing intelligence. It helps companies figure out which conferences to attend by tying together where your prospects will actually be. The bet: as AI floods digital, in-person gets more valuable, but only if you know the prospect you want is going to be in the room.
That’s an interesting bet in light of Cvent paying more than $300M for Goldcast a while back. Our prediction: Vendelux gets acquired by Cvent for around half a billion dollars before they reach a Series C.
The Feature Deal: Vista Wants to Take Criteo Private
Vista Equity Partners, alongside hedge fund Quenti Capital, has proposed taking Criteo private. The board hasn’t responded publicly yet, so this is just what we know. They offered a 50%+ premium to where the stock traded before the news, roughly $28-30 a share. The stock jumped about 20% on the report to around $23, putting Criteo near a $1.2 billion market cap, which is the value of the offer net of cash.
The context that matters: Criteo had filed to re-domicile from France to Luxembourg, specifically to make a US take-private legal. That was announced in October 2025 and completes in Q3 2026. In other words, Criteo built itself an escape hatch before any of this became public.
The revenue debate worth understanding
Two credible voices on LinkedIn framed the situation differently, and the gap between them is the whole story. Chris Sheldon noted that Criteo’s retail media revenue was reported down 32% last quarter, and that PE bid a 50% premium regardless. Ken Kubec, a former operator turned banker now at FE International, countered that per Criteo’s own filings, the decline was mostly optical, driven by an accounting change plus two large clients rolling off (Target’s Roundel and Uber Eats). Strip those out, and the underlying retail media business actually grew about 24%.
That distinction is critical to understanding why a buyer would pay a 50% premium on what looks, on the surface, like a declining asset.
For context: Criteo is a Paris and New York ad tech company with a few thousand employees and a newer CEO who came in February 2025. Vista is a large enterprise software PE firm (PitchBook currently shows ~$103B AUM with ~$11B in dry powder, and a track record cited around $350B in deals over time).
The Operator’s Read
Framed through four dimensions: strategic value, deal price, comms strategy, and post-merger integration risk.
Strategic value. Borrowing Ken Kubec’s thoughts: value in ad tech is migrating off the impression, off the demand and supply side, toward two things, first-party data and the demand model on top. Criteo sits on both, with closed-loop purchase data for 235 retailers (with an asterisk) and its new self-service platform, “Go”, pitched against Google’s PMax and Meta’s Advantage+. AppLovin proves the market will pay a premium for the data and the model, together.
But here’s the question: is Criteo in the middle or at the center? Meaning, are they a genuine value-add, or a tax? Of the ~235 connections Criteo offers, only a fraction carry meaningful volume. Drawing on Christian’s ChannelAdvisor experience (which had ~1,000 marketplace connections but where only the top 10-15 drove most of the volume), Christian argues the top retailers are more of a tax that is often not fully appreciated by brands because its easy enough to built to those channels AND most brands don’t want or need all of them. The real potential value sits in the long tail. The challenge with the long tail is lower ad volume and less platform sophistication, which impairs Criteo’s ability to feed value-added data back to advertisers. That’s precisely where a strategic acquirer’s opportunity could lie: putting Criteo in a position to do more with the network it’s already built.
Deal price. Based on the numbers, this looks like roughly $400M in profit, making it about a 2x EBITDA deal. Vista’s usual problem entering software companies is paying 10-20x and needing everything to go perfectly. Here it’s a couple of turns of EBITDA on a business throwing off ~$200M in cash with fixable inefficiencies. That’s a cheap offer for the franchise. Even with Criteo’s challenges, this asset should arguably trade at three to four times EBITDA based on current market conditions, not 2x. .
Comms Strategy. It’s messy because this is a public company. Lawyers for Vista, the hedge fund, and Criteo have almost certainly been talking behind the scenes, but Vista chose to go public with the offer, likely to pressure the Criteo larger investors to push the board to the table, or possibly draw out a competitive bid that turns this into an auction where Vista or any other buyer pays a price the minimizes, but does not eliminate the risk of shareholder lawsuits. Either way, the lack of any public response so far is telling.
Post-merger integration risk. Criteo needs product oxygen and operating efficiency. But this is a French-based firm, and as we discussed with the Vibe.co deal, “operating efficiency” on large French headcounts is not easy or cheap to execute. Criteo needs not just product innovation but probably some M&A, and Vista is good at the latter, less so at doubling down on innovation. The risk: the Vista playbook isn’t an exciting story for the ~900 engineers in the business.
The Deal Architect’s Read
Ayelet took the view of incentives and the people behind the behaviors.
Start with the structure of the business: Criteo has two sides, retail media and retargeting. The market is treating the whole thing like it’s dying, but the side that actually matters is growing. Once you understand that, the incentives make sense.
What does Vista want? Something the market has mispriced that they can buy at a ‘reasonable’ price, take out of the public-market pain, and use to own an agentic commerce option. Criteo’s current AI integration with ChatGPT is real but half-baked, exactly the kind of thing a focused owner could sharpen.
What does Criteo want? Its shareholders want to stop watching the stock drop. The company is cash-rich and has weathered tough positions, but it’s been optically punished, so a good outcome in the public market was unlikely.
So the real gap isn’t between buyer and seller. It’s between what the public sees and the reality of the business. Criteo needed a new place to swim without the weights on, and whether that’s a pool or an ocean, it needed out of the current tank.
The unfair part, as Christian put it, is that a public company living quarter to quarter off investor calls often can’t do the housekeeping it needs to do. Criteo’s 52-week high was around $26 and it was swimming around $15 before the news; the pop to ~$22.76 shows there are still fishing weights hanging off the stock even amid the speculation.
The simplest way to sum it up: Criteo is an example, not the exception. As Ken Kubec put it, there’s a long list of orphaned, profitable, cash-generating software and data companies stranded below their intrinsic value because the market narrative soured. Criteo is PE target number X. DoubleVerify, PubMatic, Magnite, PE is coming for you.
This show isn’t over; we could easily see this deal get traction around the three-to-four-times range with another bidder at the table.
Quick Hits: Two Deals, Two Opposite Structures
Descartes Systems Group acquires Driven. One of Christian’s favorite acquirers, Descartes (NASDAQ/TSX) bought Driven, a Chile-based AI route optimization, last-mile delivery, and fleet telematics platform serving Latin America. The rationale: extend Descartes’ routing and last-mile capabilities into LatAm. The structure is up to $35M, $30M cash up front plus $5M in potential performance earnouts. What’s great about Descartes is they always publish the price no matter the size, which gives the market a read on how public buyers value smaller businesses. We don’t have Driven’s metrics, but it’s likely a low multiple.
Banzai acquires ConnectAndSell. Banzai, a small public martech co with about 150,000 customers (reportedly including Amazon), bought ConnectAndSell, an AI sales tool, for $13.2M, more than 3x Banzai’s market cap, roughly doubling revenue at an 86% margin. The structure is interesting: $5.5M was paid in cash; the rest was seller note and stock, so nearly 60% of the deal is paper. The seller took most of it on paper, betting on the upside rather than cashing out.
ConnectAndSell has roughly 50 employees in the US and 20 in India, and this reads like the kind of deal where both companies needed it, a “better-together” that could practically have been a merger even though it wasn’t publicized as one. It’s the exact structure we talk about constantly: when two businesses need each other about equally, you can build a deal that’s attractive to both sets of investors and founders, with the upside sitting in the combined capital appreciation.
Other Deals we did not cover on the live show:
Miroma Group x Ad Results Media: London’s Miroma takes a controlling interest in the top US podcast/audio ad agency, buying control from PE firm Shamrock Capital (the one you keep joking about inviting on the show), which keeps a minority. Expands Miroma’s audio/podcast reach.
Chris Erwin did a great write up on this deal, here.
Accenture x Mjølner Informatics: Accenture adds a Danish digital-engineering/software group, deepening its Nordic product-engineering bench. (Accenture “at it again” — cadence beat.)
Lumine Group x Imagine Communications: a Constellation Software–style permanent-capital serial acquirer buys broadcast + AI ad-monetization tech and runs it autonomously. (This was my recommended quick hit for feature-fit — patient capital consolidating media tech.)
FINN Partners x Honner: FINN acquires a 25-person Sydney financial & corporate communications agency, opening its Australia presence.
Havas x MUT: Havas picks up a Barcelona experiential-marketing agency, extending its Spain/experiential capabilities.
That’s all for this week, have a great weekend!
Ayelet & Christian
