Mobile app advertising
You are not buying installs. You are buying the chance to ask for one, and on iOS a measurable share of those chances never arrive.
9 August 2026 ยท PAD team
The three ways to buy
Almost all mobile app advertising money moves through three routes, and they behave differently at the moment of the tap.
| Route | What you pay for | Where the tap lands |
|---|---|---|
| App install campaigns | installs, optimised | store sheet inside the app, no browser |
| Traffic campaigns | clicks | the platform's own in-app browser |
| Creator and organic posts | reach or a fee | the platform's own in-app browser |
That third column is the whole article. Install campaigns hand the user to a native store sheet, which is why they rarely show this failure. The other two go through a browser the platform wrote, and that browser cannot always complete the handoff to the App Store.
What a click means on each platform
Every ads manager counts a click when the tap leaves the ad. None of them counts whether the destination loaded. That gap is invisible in the dashboard and expensive in the bank account.
Measured on 21,000 taps routed between 3 and 9 August 2026, Facebook placements needed a manual fallback button around 1.7% of the time, against 0.05% on Instagram, with the same routing code on both. Without routing, both numbers would be the share of taps that simply stopped.
What advertising costs now
Reported averages put iOS cost per install near six dollars and Android near two, with wide variance by category: casual games at the low end, finance and subscription apps several times higher. Published benchmarks such as those from Business of Apps are useful for order of magnitude and useless for your specific case.
The number worth computing is your effective cost per install, which is spend divided by installs that actually happened, not by clicks the platform charged you for. If a fifth of taps never arrive, your real cost is a quarter higher than the dashboard says.
Objective choice changes the physics
This is the single largest lever most advertisers never pull deliberately.
| If you run | Delivery risk | Attribution |
|---|---|---|
| App install objective | low | SKAdNetwork, coarse |
| Traffic objective to a store URL | high on iOS | your own, if the link survives |
| Traffic objective to a routing link | low | your own, per source |
Teams pick the traffic objective for good reasons: better attribution, cheaper clicks, the ability to send people to a landing page first. Those reasons are sound. The cost is that you now own the delivery problem, and most do not know they have taken it on.
How to audit a live account in twenty minutes
- Pull click to install rate per campaign, split by objective.
- If traffic campaigns sit far below install campaigns with the same creative, that gap is mechanical, not audience quality.
- Open one live ad on an iPhone, from inside the app, not from a desktop preview.
- If you see a white screen, every click in that campaign has been paying for one.
That last step is the only test that counts. A link that works when you paste it into Safari tells you nothing about how it behaves where your buyers actually tap it.
What to change
Keep the objective that suits your attribution needs, and make the destination survive an in-app browser. Those are separate decisions and most teams collapse them into one, then conclude that traffic campaigns perform badly.
Paste an ad destination into the checker and see what an iPhone actually gets.