Imagine buying a ticket to a concert, then discovering that the ticket seller counted an empty seat as a satisfied listener. Digital advertising has its own version of that problem. A server can deliver an ad without a person having a useful chance to see it. Integral Ad Science makes its living investigating what happened between the delivery receipt and the audience.
- IAS checks whether ads reach people, appear in view and run in suitable surroundings.
- Its tools help buyers filter inventory before bidding and inspect campaigns afterward.
- The useful test: does better media quality improve the economics of your campaign?
The receipt is not the result
The business began in New York in 2009 as AdSafe Media. Early co-founders included Helene Monat and Will Luttrell. The name described a pressing concern: advertisers wanted control over the company their ads kept. In 2012, AdSafe became Integral Ad Science. Its expanding remit included another uncomfortable possibility: a perfectly respectable page could carry an ad that nobody could see.
Today, IAS separates several questions that a delivery count bundles together. Was the visitor human? Was the placement viewable? Was the surrounding content appropriate for this particular advertiser? Was it in the intended geography? Its Quality Impressions metric brings those conditions together. That provides a more demanding denominator for judging a buy. Passing the checks establishes an opportunity for advertising to work; it does not establish that someone paid attention or purchased anything.
Suitability makes the problem more interesting than a simple blacklist. Consider a hypothetical wine brand and a children’s brand looking at the same article about alcohol. Their requirements could differ even when the page is legitimate. IAS uses contextual analysis, including sentiment, to help distinguish environments. The buyer still has to decide which associations fit the campaign.
The expensive moment is before the bid
An inspection after purchase helps explain waste. An inspection before purchase can help prevent it. IAS supplies pre-bid filters inside buying platforms, alongside post-bid measurement and blocking. Quality Sync aligns settings across those stages. Context Control helps select relevant environments or avoid unsuitable ones. IAS Signal presents campaign reporting. Quality Attention adds an estimate of attention using media signals, machine learning and Lumen Research’s eye-tracking data.
The company sells these capabilities to brands and agencies, while also serving publishers and platforms. Fees commonly follow impression volume, expressed as a cost per thousand impressions, or CPM. Google’s buying-platform documentation makes a practical point: selecting different types of quality controls can add their fees together. A buyer should compare the complete bill with the value of the inventory that survives.
A $2 CPM with half the impressions usable means $4 per thousand usable impressions, before verification fees. Illustrative arithmetic, not IAS pricing or a campaign result.
Mitsubishi puts the premise on the road
For the 2026 Outlander launch, Mitsubishi Canada tested an automated approach to media-quality targeting with IAS. In an August 2026 account of the controlled study, IAS reported an 82% reduction in brand risk, a 33% increase in click-through rate, a 15% improvement in clicks-to-visit rate and a 4% lower cost per visit.
less brand risk
higher click-through rate
lower cost per visit
The modest final percentage deserves attention. It connects the quality discussion to a commercial cost. These are company-reported findings from one test, with no guarantee that another advertiser will reproduce them. The useful thing to copy is the experiment: change the buying rules, then inspect both placement quality and the cost of a meaningful response.
The publisher needs the rulebook, too
There is a less glamorous source of waste: publishers trying to satisfy campaign requirements they cannot fully see. IAS addressed that information gap with Quality Connect in June 2026. When advertisers consent, publishers can inspect their quality preferences inside IAS Pulse. The initial Campaign Transparency capability launched at no additional cost. The Financial Times described easier campaign setup and less back-and-forth.
“The quality of a media buy shouldn’t be a secret.”
Lisa Utzschneider · Quality Connect announcement, June 2026
That product choice fits the company’s stated emphasis on accountability and customer focus. The harder operational lesson concerns incentives: an advertiser’s protective setting can become a publisher’s blocked impression. Showing both sides the same requirements gives them something concrete to fix.
The screen keeps changing
IAS’s expansion into streaming required more than extending a browser report. In August 2021, it acquired Publica for $220 million in cash and stock. Publica brought connected-TV ad serving and unified auctions for publishers. The purchase put IAS closer to the machinery that monetizes streaming inventory.
In April 2026, IAS introduced Total TV, combining quality and outcome information with aggregated content insights from participating providers including Disney, NBCUniversal, Paramount and Prime Video. In June, episode-level podcast controls arrived through Spotify Ad Exchange and The Trade Desk. The distinction matters: one podcast episode can present a different context from the next.
Ownership changed along the way. IAS listed on Nasdaq in 2021; Novacap completed its approximately $1.9 billion take-private acquisition in December 2025. Lidiane Jones, previously CEO of Slack and Bumble, succeeded Lisa Utzschneider in July 2026. In October, IAS announced a closed pilot for brand safety and suitability measurement in ChatGPT Ads. Each new format introduces another environment that buyers want independently examined.
Buy a better question
DoubleVerify offers a broad alternative; fraud specialists such as HUMAN address narrower parts of the problem. IAS’s pitch combines contextual analysis, fraud expertise, measurement, buying-platform integrations and publisher tools. The attraction is the ability to move from an observed problem to a buying decision within connected workflows.
For a working marketing team, that suggests a specific brief: identify a placement problem, choose the relevant control and agree on the result that would justify its fee. Keep the campaign’s commercial objective in view. Removing unsuitable inventory can be worthwhile, but a cleaner report alone cannot tell you whether the offer, creative or landing page persuaded anyone.
Those workflows have boundaries. Google documents that unknown URLs can be excluded when classification information is unavailable. Strict rules can therefore cost reach. Streaming transparency depends on participating publishers, and available controls differ by channel. A sensible buyer starts with a defined problem, checks platform coverage, aligns settings and tests against a comparison campaign. The question worth paying to answer is wonderfully ordinary: did the next dollar buy a better opportunity?