If you run Display or Performance Max campaigns, a share of your budget is almost certainly being spent inside mobile apps and games — and a large proportion of clicks from that inventory are accidental taps rather than interest. This is not fraud in the criminal sense. It costs you exactly the same, it is switched on by default, and unlike most of the problems in this cluster it has a complete fix that takes about five minutes.
Why app inventory behaves differently
Picture the actual moment. Someone is playing a game on a phone. A banner appears at the edge of the screen, often near a button they were already reaching for. They tap it by mistake, your page begins to load, they hit back immediately.
You paid for that click. From Google's side it is a legitimate click on a legitimate placement — no bot was involved and nobody defrauded anybody. From your side it is indistinguishable from waste.
Two structural features make this worse than ordinary poor targeting. Ad units in games are frequently placed where thumbs already are, which produces a click-through rate that looks excellent and converts at close to zero. And a specific category of app — usually free, ad-heavy children's games — generates enormous impression volume with essentially no commercial intent behind any of it.
Checking what it is costing you
For Display campaigns, the placement report tells you directly. Open the campaign and go to Content → Where ads showed (in some account layouts this is under Placements). Sort by cost, descending.
What you are looking for is app placements — they appear with names like
mobileapp::2-com.example.somegame — sitting high in your spend with conversion counts
at or near zero. If several of your top ten cost lines are games, you have found real money.
Performance Max is harder, deliberately so: Google reports far less placement detail and gives you far fewer controls. You can see some placement data, but you cannot manage it with the same precision. That limitation is worth knowing before you assume a Performance Max campaign can be cleaned up the same way.
Check conversions, not click-through rate. App placements often show a flattering CTR precisely because the taps are accidental. Judging this inventory on CTR will tell you it is your best-performing placement. Judge it on cost per conversion, and ideally on cost per actual customer.
The three exclusions worth making
1. Exclude app categories wholesale
At account level, under Content suitability or your campaign's placement exclusions, you can exclude broad app categories. For most B2B and considered-purchase businesses, excluding games outright is a decision you will not regret — the audience overlap is small and the accidental-tap rate is the highest of any inventory type.
2. Exclude all app inventory, if apps are not your market
There is a placement exclusion that removes mobile app inventory entirely. Add
mobileappcategory::69500 as a placement exclusion at campaign or account level and your ads
stop serving in apps altogether. If you are a local service business, a clinic, a school or a B2B
supplier, this is very often the correct setting and almost nobody has it applied.
3. Exclude individual offenders as you find them
For anything left, add specific placements to your exclusion list as they surface in the report. This is ongoing rather than one-off, which is why the two blanket exclusions above are the better use of your time if apps are simply not where your customers are.
What to expect afterwards
Impressions and clicks will fall, often sharply. Cost per click will usually rise, because you have removed the cheapest inventory you had. Neither of those is a problem, and both will look alarming on a dashboard for about a week.
The number that matters is total conversions, and specifically total conversions that turned into real conversations. If that holds steady while spend drops, the exclusions worked exactly as intended. If it falls proportionally, apps were genuinely contributing and you should reconsider — that is uncommon for service businesses and entirely plausible for a consumer app or a mobile-first product.
Why this keeps happening
Google Ads defaults expand reach. That is a reasonable default for an advertiser who wants volume, and a poor one for a business with a narrow, high-value audience — which describes most of the companies reading this.
Auditing every default forever is not a realistic strategy. The durable version is to make sure the conversions your bidding optimises toward are real customers rather than form fills, because then a placement producing taps and no revenue stops attracting budget on its own, without anyone finding it first. That is what offline conversion imports are for, and what Claudphic Ads reports back automatically when a lead is marked Won. Exclude the games today; fix the feedback loop so the next bad placement handles itself.