Keyword Count Per Ad Group in Apple Search Ads: How Big Is Too Big for Bid Control
Most indie developers build their first Apple Search Ads campaign the same way: open a Search Results campaign, paste in every keyword they can think of, and let the bids sort it out. It feels efficient. One ad group, forty terms, done. A few weeks later the numbers look muddy, a handful of keywords are eating budget, and nobody can say why a term with a good tap rate never converted.
The size of your ad groups is one of the least discussed levers in Apple Search Ads, and it quietly shapes how much control you really have. This post is about how to think about keyword count per ad group, when a crowded group starts hurting you, and how to decide where to split.
Why group size matters at all
In Apple Search Ads, the bid lives on the keyword inside an ad group. An ad group also holds settings that apply to everything inside it, such as the match type mix, the default bid behavior you set, and the Search Match toggle. Your reporting, however, only gets sharp at the keyword and search term level. So the ad group is the unit where you make structural decisions, and the keyword is the unit where you make bid decisions.
When an ad group holds too many keywords that behave differently, you end up making a single compromise. You set a bid that is fair to the best term, which overpays for the weak ones, or you set a bid that suits the weak terms, which starves the strong ones. Either way the group's average hides what is happening underneath.
The signs your group is too crowded
You do not need a magic number to notice a problem. Look for these patterns:
- Wildly different tap-through rates sit side by side. One keyword gets taps on most of its impressions while another, in the same group and with a similar bid, barely gets any. That gap is worth investigating, but it is hard to act on when the terms share a structure you cannot change independently.
- Your best keyword stops getting impressions after you touch a weak one. Raising or lowering a bid in a group can shift delivery for neighbors in ways that are hard to attribute, especially if they compete for the same searches.
- You cannot describe what the group is for in one sentence. If the honest answer is "it's all the keywords for our app," the group is probably doing too many jobs.
- You keep adding terms but never remove any. A growing list with no pruning usually means no one is confident enough to judge the older keywords against the new ones.
None of these is proof on its own. Together, they suggest the structure is making decisions harder than they need to be.
A practical way to decide group boundaries
Instead of asking "how many keywords is ideal," ask what a single bid decision should cover. Group keywords that share three things:
1. Similar intent
A searcher typing a specific feature name, a category term, and a competitor's name is in three different mental states. Their willingness to convert differs, and so does the bid that makes sense. Keep terms with similar intent together so one bid range fits the whole group.
2. Similar expected value
If one keyword is something you would happily pay a higher CPT for and another is a curiosity, separate them. You want the high-value group to be able to hold a bid the low-value group should never have.
3. Similar review rhythm
Some keywords need daily attention while you are testing them. Others you check once a week. Grouping by how often you will look at them keeps your routine sane and keeps you from fiddling with stable terms because a volatile one is nearby.
A useful rule of thumb is that a group should be small enough that you can read its keyword table in one glance and explain each row. The exact count depends on your app and budget, so treat any specific number as a starting guess rather than a rule.
Illustrative example
Suppose a fitness tracker app has one ad group with thirty keywords. It mixes brand-adjacent terms, generic category terms, and a few specific feature phrases. Over two weeks, the feature phrases convert well but get modest impressions, the category terms get large impressions with weak conversion, and the group bid sits in the middle.
A cleaner structure would split the feature phrases into their own ad group with a bid tuned to their intent, and move the broad category terms into a separate group where you can test a lower bid and watch conversion closely. Nothing about the keywords changed. Only your ability to judge and adjust each one did. This is an illustration of the reasoning, not a promise about any particular result.
What to do this week
- Export your keyword table and sort it by impressions, taps, installs, and conversion rate, one ad group at a time.
- For each group, write one sentence describing its purpose. If you cannot, mark it for splitting.
- Look for keywords whose intent does not match the group's sentence and move them out.
- Pause nothing yet. Restructure first, then give each new group a few days before judging it, so you are not reacting to the disruption itself.
- Keep a short note of what you moved and why. Future you will want to know which structure produced which numbers.
Closing takeaway
Keyword count per ad group is not a score to maximize. It is a question of whether you can still make a clear decision about each bid. Keep groups focused on one intent and one value level, check that you can explain every row, and split when a single bid can no longer serve everything inside it.
If you find this kind of review tedious across a growing set of campaigns, AdsBuddy is built to surface these structural issues as a short, prioritized list each day, and you still approve every change yourself.