← All posts
November 12, 2026

Tap-to-Install Rate as a Product Page Diagnostic: Is It the Keyword or the Page?

Apple Search AdsASOproduct pageconversion rateindie iOSCPTROAS

Every indie dev who runs Apple Search Ads eventually hits the same frustrating row in the report. A keyword has a healthy tap-through rate, the cost per tap looks reasonable, and yet the installs barely show up. The instinct is to blame the keyword and cut the bid. Sometimes that is right. Often it is the wrong move, because the problem lives on the product page, not in the auction.

This post walks through how to separate those two causes using one metric you already have: conversion rate from tap to install. Once you know how to read it, you stop guessing and start deciding where to spend your attention.

What the tap-to-install rate actually measures

In Apple Search Ads, the conversion rate you care about for a keyword is installs divided by taps. It answers a narrow question: of the people who looked at your ad and chose to tap it, how many went on to install? It says nothing about whether those people were a good fit for your app in the long run. That is a revenue question, and you answer it later, by following the install to purchase chain.

The key point is that by the time someone taps, Apple has already done the matching work. The query they typed, the keyword you bid on, and the ad they saw are all behind them. What is left between the tap and the install is your product page: the icon, screenshots, preview video, subtitle, and description, plus whatever custom product page the ad group routes to.

That means a low tap-to-install rate tells you something about the page for that traffic. A healthy one tells you the page is doing its job for the people who arrived.

Two symptoms that look the same but aren't

The report can show you a weak result in two different shapes, and they call for different responses.

Symptom one: taps come in, installs do not

Here the tap-through rate is fine and the cost per tap is in an acceptable range, but installs per tap are low compared with the rest of your account. Possible causes:

  • The query is loosely related to what your app does, so the searcher tapped out of curiosity and bounced once they saw the listing.
  • The page does not speak to the specific promise the query implies. A searcher who typed a narrow task phrase lands on a generic screenshot set and leaves.
  • The page has a conversion problem for everyone, and this keyword just exposes it more clearly because it sends a larger volume of traffic.

Symptom two: installs look normal, but the keyword is still expensive

This one is sneakier. Tap-to-install is healthy, so the page is fine. The cost per install is high anyway because the cost per tap is high. The fix here is on the bid side, or on the decision to keep the keyword at all. You are paying for a competitive auction slot, and the page is not the bottleneck.

If you only look at cost per install, these two cases blur together. Separating taps, installs, and CPT as distinct numbers keeps them apart.

A simple diagnostic sequence

When a keyword looks weak, work through the checks in this order before you touch its bid:

  1. Confirm sample size. A keyword with a handful of taps tells you very little about its conversion rate. Wait until you have enough taps that one or two installs do not swing the result wildly. If you are unsure, compare against the rest of your account instead of a fixed threshold.
  2. Compare against your baseline. Pull the tap-to-install rate for your whole Search Results campaign, then for the individual ad group. If this keyword sits far below both, you have a signal worth investigating. If it sits near the baseline, the page is probably not the issue.
  3. Check which page it routes to. Confirm whether the ad group points at your default product page or a custom product page. A keyword sent to a generic page is a common mismatch.
  4. Read the search terms. Look at the actual queries that produced those taps. If the matched queries drift far from the keyword's meaning, the problem is query relevance, not the page.
  5. Only then decide on the bid. If the page is converting fine and the cost per tap is the issue, adjust bids. If the page is underperforming for a specific intent, change the page first.

Fixing the page for a specific intent

When the diagnosis points at the page, resist the urge to rewrite everything at once. Change one dimension, give it time to accumulate taps, and compare.

  • Match the first screenshot to the promise. If your keyword is about a particular task, the first screenshot should show that task being completed. A generic brand image wastes the strongest real estate on the page.
  • Use a custom product page for a distinct intent cluster. Route the ad group for that cluster to a page that mirrors its language. Keep the rest of your traffic on the default page so you can compare.
  • Align the subtitle and opening description with the query family. Searchers scan for their own words. If the words are missing from the top of the page, the tap is more likely to bounce.
  • Do not chase the highest conversion rate by narrowing traffic to nothing. A page that converts beautifully for a tiny trickle of taps may not be worth optimizing toward, so keep the volume question in view.

After a change, judge the page against the same kind of traffic. Comparing a new page's week of data to a month of mixed traffic gives you a false improvement or a false loss.

What this does not tell you

Tap-to-install rate is a useful first filter, but it is not the final verdict. A keyword with a modest install rate can still produce users who buy and renew, and one with a strong install rate can attract people who never pay. The revenue side of the chain, typically handled through a subscription or purchase tool that maps the attribution token to events, is what closes the loop. Use conversion rate to decide where to look, and use revenue to decide what to keep.

Also keep in mind that Apple does not hand you per-keyword revenue directly. You are building that picture from installs and downstream purchases, so treat any single keyword's revenue figure as an estimate that improves as data accumulates.

Takeaway

A weak keyword is not automatically a bad keyword. Before you cut a bid, check whether the tap was the problem or the install was. Compare tap-to-install against your baseline, confirm which page the traffic reaches, read the actual queries, and only then change something. Most of the time the cheapest and most informative fix is on the product page for a specific intent, not on the auction.

If you find this kind of daily triage tedious, AdsBuddy is built to surface these checks as a short, prioritized list each day, and you still approve every change yourself before anything happens.

Run Apple Ads with AdsBuddy

Connect Apple Ads and RevenueCat, then get a short prioritized list of changes to review and apply. Plans from $29/mo.

Get started
Get new posts in your inbox
Practical Apple Search Ads tactics for indie iOS devs. No spam, unsubscribe anytime.