The comparison comes up a lot in boardrooms right now, with many asking if the current AI boom is just ad tech’s version of the late-90s dotcom bubble? Are we in the “Flooz.com” phase — or already drifting toward the shake out?
If you look at 2025’s deal tape, it’s easy to see why people ask. The year began with genuine froth: a more business-friendly U.S. administration, falling-rate expectations, and early trophy prints such as T-Mobile’s double-swoop on Vistar Media (for approximately $600 million) and Blis ($175 million), plus strategic moves like Publicis buying Lotame and The Trade Desk picking up Sincera.
But by Q3, dealmakers were talking about something closer to a controlled deflation than a mania. Global mergers and acquisition volumes across talent- and tech-enabled services are down about 8% year on year, with buyers citing macro volatility and a widening valuation gap as the main reasons processes stall, according to sources. For many, this represents the much-touted 2025 rebound in M&A as arriving with “a whimper, not a bang,” as bankers lean into smaller, more surgical transactions instead of 2021-style land grabs.
Yet AI is in almost every pitch deck and virtually every process. The question is whether that AI story is real — or just lipstick on a pig.
One of the clearest signals from SI Global’s latest Global M&A Insights Report is that AI has shifted from “nice to have” to table stakes. Buyers say a well-developed AI strategy is now a baseline requirement, and 70% of respondents name data and analytics as a key interest for next year’s M&A; roughly half explicitly call out AI and machine learning.
Paul Silver, global president of corporate development at MiQ, describes a diet change rather than a loss of appetite. Private equity hasn’t gone off ad tech, he argues — it’s moved from “growth-at-all-costs” to “consolidation-and-integration.” The new playbook is to roll up “orphan assets” with solid tech but broken cap tables into larger platforms, strip out duplicate back-office costs, and use AI to lift margins rather than justify wild revenue multiples.
In practice, that’s showing up as smaller, highly targeted deals. Verve Group’s proposed acquisition of Captify is a neat example, costing the publicly-listed company roughly €25.6 million ($27 million) for an asset expected to contribute about €41 million ($47.7 million) of 2025 revenue and €5 million ($5.8 million) of EBITDA after synergies. It’s a classic data-plus-demand-side tuck-in: one of the largest onsite search datasets outside the walled gardens, bolted onto an existing DSP footprint rather than sold as a standalone AI “story.”
The same pattern runs through other 2025 transactions. WPP’s purchase of data-clean-room provider InfoSum, The Trade Desk’s acquisition of Sincera for a sub-$50 million sum, and DoubleVerify’s deal for Rockerbox all fall into the “defensibility buys” bucket — assets with proprietary data or infrastructure that supports AI-driven planning and measurement, rather than pure model IP.
Meanwhile, AI is also driving capability grabs at the creative and SMB end of the market: Magnite’s acquisition of streamr.ai to simplify CTV creative and campaign setup for small advertisers, and Rembrand’s merger with Spaceback to fuse virtual product placement with social-to-programmatic creative automation. Both speak to a barbell that Silver and others see emerging: CTV scale on one end, AI-enabled margin expansion on the other.
Paul Knegten, an industry consultant with extensive experience working with ad tech companies, described a “huge force from AI” that’s forcing many investors to reevaluate any earlier theses — and could prompt M&A activity.
“On one hand, the pressure’s gotten even stronger to consolidate that’s just, there’s this massive, like existential pressure,” he said, describing this as “existential.” Knegten added, “You have pure-play AI companies that are innovating directly in that space. Then you have, like, the rest of the ad tech infrastructure that needs to kind of get consolidated quickly into some very durable companies.”
PE’s patience meets the AI overhang
The more uncomfortable dotcom parallel sits inside PE portfolios. SI Global’s research shows that 21% of PE-backed B2B assets are now classed as “overdue,” double last year’s level, while refinancings are down 57%. In plain English: a big chunk of the 2020–21 vintag