Generic Category Keywords in Apple Search Ads: Why Head Terms Look Cheap on Paper and Expensive in Installs
Every Apple Search Ads account eventually grows a few short, generic terms. Things like 'budget app', 'habit tracker', or 'photo editor'. They show up with big impression counts, they feel like the obvious place to be, and they often look cheap per tap. The trouble is that a generic term rarely means one kind of searcher. It means a crowd, and the crowd is usually a mix of people who want exactly what you sell and people who want something adjacent. This post walks through how to tell those apart using only the numbers Apple gives you, and how to set bids on category terms without fooling yourself.
Why head terms look better than they are
A short category term has three properties that flatter it in a quick glance:
- High impressions. Many people search for it, so the volume numbers look healthy.
- Decent-looking TTR. Users who type a broad phrase often tap something on the results page, so taps look reasonable.
- Low-ish CPT relative to competitive brand terms. Plenty of apps bid on the same generic phrase, and the auction doesn't always push the price up, so your actual cost per tap can look friendly.
None of those three properties tells you whether the tap turns into an install, and none of them tells you whether the install turns into revenue. A cheap tap that goes nowhere is still a cost. Your job is to follow the chain past the tap.
Measure the chain, not the top of the funnel
For a category term, read the metrics in this order:
- Impressions to confirm the term actually serves for your storefront.
- TTR (taps divided by impressions) to see whether the result is attracting clicks at all.
- Conversion rate (installs divided by taps) to see whether the people who tap are the people who install.
- CPI (spend divided by installs) to see what each install costs you.
- ROAS once you have enough time for purchases to arrive, using revenue from an attribution source such as RevenueCat, since Apple does not report revenue per keyword.
The step that most often gets skipped is step 3. A generic term can post a healthy TTR and still convert poorly, because the searcher wanted a different category of app, or wanted a free tool when your app is paid. Low conversion on a head term is a signal about intent, not necessarily about your product page.
An illustrative example
Suppose a term gets 1,000 impressions, 60 taps, and 3 installs in a week. That is a 6% TTR and a 5% conversion rate. Now suppose a more specific phrase gets 150 impressions, 30 taps, and 9 installs. The specific phrase has a lower impression count and a higher TTR, but it also has a conversion rate of 30%. The generic term looks bigger, but the specific term is producing three times the installs from a fifth of the impressions. These numbers are made up for illustration, but the pattern is common: volume and efficiency often point in opposite directions.
Separate the generic term from its specific cousins
A practical habit is to keep generic category terms in their own ad group rather than mixing them with specific terms that share a theme. Two reasons:
- Bids stay honest. If a head term and a long-tail term sit in one ad group under one bid, you can't tell which one is paying for the installs. Splitting them lets you set a max CPT on the generic group that reflects what those installs are worth, and set a different bid on the specific group.
- Reports stay readable. When the generic group's CPI climbs, you can see it immediately without the specific terms masking the change.
Keep the number of keywords in any one ad group manageable, so that a bid change on a category term is actually a change to one decision.
Set the bid from the value of an install
Before you touch a generic term's bid, write down what an install from that term is worth to you. Start from your own numbers rather than anyone else's:
- Average revenue per paying user, over a window long enough for purchases to mature.
- Share of installs that ever pay.
- Whatever margin you can keep after fees and any other costs.
Multiply those and you have a rough ceiling on what a generic install can cost before it stops making sense. If the generic term's observed CPI sits above that ceiling for several weeks, lowering the max CPT or pausing the group is the honest move. If it sits comfortably below, it may be worth holding or testing a modest increase, one change at a time.
Remember that max CPT is a ceiling, not a price. You will often pay less than your bid, so the bid tells you how much you are willing to pay, while the CPI tells you what you actually paid.
Use negatives to trim the crowd, not to kill the term
If a generic term is worth keeping but its search results are full of irrelevant queries, the right tool is usually a targeted negative keyword for the obviously wrong intent, not a blanket pause. Review the search term report, find the recurring queries that never convert, and block those specifically. Leave the core term alone so that the traffic you want keeps flowing.
When to stop bidding on a category term
Pause or retire a generic term when all of these are true over a window you have agreed with yourself in advance:
- CPI is above the value ceiling you calculated.
- Conversion rate is clearly below your specific terms in the same category.
- Negative keywords have been tried and the waste persists.
If you retire it, note the date and the reason, so that a future review doesn't reopen the question with no context.
Takeaway
Generic category terms are not bad. They are just unclear. Judge them by installs and, eventually, revenue rather than by impressions and cheap taps, keep them in their own ad groups so their bids are honest, and set their ceilings from what an install is actually worth to you. A quick, repeatable review of this chain is the kind of thing a tool like AdsBuddy can surface as a short daily list, while you still decide and apply each change yourself.