Who Verifies the Verifiers?

In late 2025, Integral Ad Science exited public markets, sold to Canadian private equity firm Novacap for roughly $1.9bn, announced in September and completed in December. In early August 2026, its larger rival DoubleVerify agreed to a $2.15bn take-private deal with Nielsen, an all-cash transaction at a roughly 30% premium, expected to close by Q1 2027. Two companies whose entire commercial promise rested on being neutral outside referees, checking whether a brand's media dollars landed on real, safe, viewable inventory, both exited independent ownership within eight months of each other. One went to a legacy measurement company that also sells advertisers other things. The other went to a financial sponsor whose job is maximising exit value, not maximising scrutiny.

The timing sharpens the question. DoubleVerify's own research, published in July, found that poor-quality AI-generated content is actively putting brand trust at risk, an admission from inside the industry that the supply chain it is paid to police is getting murkier as generative tools flood the open web. In the same window, DV spent June extending its pre-bid brand safety product into Meta and TikTok, platforms that have historically resisted letting outside measurement vendors see too much. That is a genuinely useful expansion for advertisers. It is also DV becoming more entangled with the platforms it is meant to be checking, not less.

The evidence of harm is real and specific, not a vague industry anxiety. Lunio's State of Click Fraud Report 2026, surveying 131 senior marketers in May and released in July, found 75.6% estimate losing more than 5% of monthly performance budget to invalid traffic. Separately, the Affinity Solutions Outcomes Marketing Council surveyed 210 senior brand and agency marketers in March 2026 and found 91% believe platform-reported results are overstated to some degree, and more than two-thirds estimate at least 11% of their media budget is wasted to optimisation lag, more than a third put the figure above 26%.

Which makes the next number the most uncomfortable one here: in the same Lunio survey, only 5.3% of marketers run a dedicated invalid-traffic tool. Lunio itself calls this the 95% execution gap. Three in four know they are losing money. One in twenty is actually paying to stop it.

There is real evidence the gap is irrational, not just uncomfortable. DoubleVerify reported on 29 July that fraud and invalid-traffic violation rates fell 41% year over year in North America and 45% in EMEA among campaigns actually running its protection. Verification demonstrably works when it is used. The gap is not a question of whether it works, it is a question of who inside the client organisation feels the cost of skipping it.

The honest read is not that independence is dead, both deals are explicable on ordinary commercial terms, and neither is scandalous on its own. But the question the industry has mostly avoided asking out loud is what independent verification even means once the verifiers are owned by the kind of entities they are supposed to be checking up on.

Price alone does not explain a 75.6-to-5.3 gap this wide. Verification is typically billed as a percentage on top of media spend, so it shows up as a visible line item exactly when budgets are under the most pressure to shrink. But the deeper explanation is a split incentive: the budget line for verification is usually approved by a procurement or marketing-ops function optimising for cost efficiency this quarter, while the cost of skipping it, a brand safety incident, wasted spend, bad data feeding automated bidding, lands with the CMO's credibility or a finance team trying to explain underperformance. When the person who signs the check and the person who feels the consequence are not the same person, the check tends not to get signed.

There is also a genuinely awkward incentive sitting inside the agency itself. Media trading has a long, well-documented history of margin sitting in places clients cannot easily see, brought into sharp focus by the ANA's 2016 transparency investigation and never fully resolved since. An agency compensated as a percentage of media spend has a soft, rarely-spoken incentive to keep that spend number large. Invalid traffic filtering, done properly, shrinks the pool of impressions a client ends up paying for, exactly the outcome the client wants and exactly the outcome that can work against the agency's own commercial model. The agencies that sell verification well tend to be the ones whose fee structure does not punish them for reducing the client's spend.

Not every player in this category is heading toward consolidation. Pixability has stayed independent and deliberately narrow, the only partner Google has certified specifically for YouTube brand suitability, and in 2024 it paired its own technology with Ad Fontes Media's news-quality ratings rather than building a general-purpose tool to compete head-on with DV and IAS. Being too useful and too specific to easily replace is its own form of protection in a category where the two biggest names just proved that scale attracts acquirers.

Ask directly, inside your own organisation, who actually owns the consequence of a verification gap, and check whether that person is the same one who approves the line item. If they are not the same person, that mismatch, not the price of the tool, is very likely the real reason coverage is thinner than the risk. Separately: if your agency bundles verification into the default plan rather than pitching it as an optional add-on, ask why, and ask what your agency's own fee structure would look like if invalid traffic filtering meaningfully shrank your media spend, that answer tells you more about their incentive to sell it properly than anything in a case study. And watch, not the acquisition price tags, but what Nielsen and Novacap actually do with DV and IAS pricing and product roadmaps now that both companies answer to owners with their own commercial interests in the media supply chain.