Native and app open ads: are they worth adding?
Two formats most apps add after rewarded, interstitial and banner. Where each one fits, where it backfires, and what Google's own guidance says.
Once the three main formats are running, most developers look at native and app open ads next. Both can work. Both can also make an app feel worse than the money is worth, so it helps to know where each one fits.
Native ads
With a native ad, the network doesn't send a finished banner. It sends the pieces (headline, image, icon, button text) and your app lays them out in its own style. A good one looks like another card in your feed, with an "Ad" label on it.
That's why they suit apps that already have feeds or lists: news, recipes, marketplaces, social apps. A sponsored card every eight or ten items fits in without much friction.
If your app doesn't have a feed, native is a lot of work for very little. You have to design the layout, handle missing images, and follow the network's rules on labelling. Google, for example, requires native ads to be clearly identifiable as ads.
App open ads
These are full-screen ads that appear while your app loads or when someone switches back to it. Google describes them as "designed to be shown when your users bring your app to the foreground", and recommends putting them on your splash or loading screen.
They fit utilities that people open many times a day: flashlights, scanners, calculators, VPNs. Each launch is a chance to show an ad, and the ad doesn't interrupt anything once the app is running.
Two places to be careful:
- The very first launch. Someone who has just installed your app should see the app, not an ad.
- Quick switches. If a user pops out to copy a code and comes straight back, an ad every time gets old fast. Google's AdMob help page on app open ads notes they perform best in apps opened more than once every four hours, and says not to show other full-screen ads right before or after one.
What do they pay?
We looked, and we couldn't find a neutral public benchmark for native or app open eCPMs like the one Tenjin and CAS publish for rewarded and interstitial ads. Vendor numbers exist, but they vary too much to quote.
So the calculator uses cautious estimates for both, and it labels them that way. Your own data is better: run the format for a slice of users for two weeks, then compare revenue per user and day-7 retention against users who didn't see it. If revenue goes up and retention holds, keep it.