Mobile advertising has a problem: in-app ad traffic is riddled with fraud.
This is according to a new Pixalate report that found that 41% of global open programmatic mobile app traffic was invalid in Q2 2026, based on its analysis of 80 billion ad impressions.
Why This Matters:
Hmm, that’s a lot of questionable traffic. It’s also much higher than what Pixalate found elsewhere: IVT was 20% across desktop and mobile web and 26% across CTV.
And the problem was fairly widespread geographically. Mobile in-app IVT ranged from 35% in LATAM to nearly half (49%) in APAC. (Yeesh!) North America came in at 39%, while EMEA was 37%.
The findings speak to a bigger challenge for in-app ad traffic and in-app advertising more broadly. Many advertisers, particularly those outside of more performance-based industries (e.g., gaming, ecommerce, etc.), have long been wary of the channel because of concerns about fraud, low-quality inventory and whether they’re getting what they paid for.
A 41% IVT rate won’t exactly help change that perception, of course.
Experts React:
Possible fraud isn’t the only thing hurting the perception of in-app advertising. The ad experience itself can sometimes feel overwhelmingly low-quality, with formats that make it difficult for users to skip or even close an ad.
In February, Jeromy Sonne, founder of Daypart AI, highlighted an ad inside a Zynga game that he said gave users no apparent way to exit other than opening the Google Play Store, forcing an interaction. That’s not necessarily fraud, but it’s not exactly a great ad experience, either.
Our Take:
This matters for companies like AppLovin and Moloco, which are trying to convince more non-gaming advertisers to spend inside mobile apps.
They need brands to view in-app inventory as a legitimate alternative to the web and other major ad channels. Numbers like this, from Pixalate, make that pitch harder. Even if the reality is more complicated than “mobile apps are filled with fraud,” that perception is an uphill battle these companies will have to overcome.
