Growth teams ask me which ad channel to add next. The map below uses numbers from one subscription app: four channels, same app, same period, day 28 gross media ROAS, most of it on $181.8K of Meta spend.
The dashed marker is 1.0x, gross media parity. TikTok's cost was not tracked in the MMP, so its ROAS is an estimate. This is one app at unequal spend levels, a map of one account and not a randomized channel comparison. By my arithmetic, the four channels add up to $207.2K of spend and Meta carried 88% of it, so the other three figures rest on $25.4K between them. Gross media parity leaves out store fees, taxes and refunds, and attributed ROAS is not incremental ROAS.
What each channel is good for
| Channel | Good for | Watch out for |
|---|---|---|
| Meta | The top of the stack, especially for iOS. It scales if you are willing to test new markets. | Creative is everything, and scaling takes work |
| Google Ads | Android. Once it works it takes way less of your time than Meta, and iOS campaigns outside the EU can hit similar quality to Android. | Tough to crack if your Android app does not make money |
| TikTok | Gen Z and bargain hunters | Even harder than Meta to figure out creatively, and rarely your main channel |
| Apple Search Ads | Solid users who search related terms | Pricey because the barrier is low, caps out fast, and bidding on your own brand will mess up your numbers |
Google announced in 2026 that it is extending its iOS measurement support to users in the EEA, UK and Switzerland, so check what your account supports before you rule out Europe. As of 29 September 2026, Google’s on device measurement page still says the feature is inactive for users in the EEA, the UK and Switzerland.
Each channel’s ROAS is in the chart above.
I set out when Google, TikTok or Apple Ads should come before Meta in a separate post, and I tested which Google Ads bid strategies and networks pay off for apps across $100K of spend.
The shiny channel trap
I see a lot of companies chasing shiny new channels without realizing they’d get way better ROI putting that same effort into Meta. And scaling Meta isn’t some easy win either, you’ve got to grind for it. I laid out how I structure, test and scale Meta after Andromeda.
Before adding a channel, ask what the same hours would earn inside the channel that already works.