How Many Keywords Belong in One Ad Group?

There is no official answer, and that is worth knowing before you spend another hour looking for one. Google does not publish a recommended number of keywords per ad group. The numbers you will find quoted, and you will find several that disagree, are practitioner judgment rather than platform guidance. The test that actually decides the count is simpler than any of them: one ad group should hold only the keywords that a single set of ads can speak to honestly. The number falls out of that. It is a result, not a rule.

Why no official number exists

Google Ads has technical limits on how many items an ad group can contain, but a technical limit is a ceiling, not a recommendation, and nobody is anywhere near it. Beyond that ceiling the platform is silent, and the silence is reasonable, because the right count genuinely depends on how tightly your keywords cluster around one idea.

This is why the advice you find contradicts itself. One guide says under thirty, another says ten to twenty, another says three to ten. None of them is citing Google, because there is nothing to cite. They are each describing the point at which, in their experience, an ad group stops holding together. That is a legitimate thing to have an opinion about. It is just not a number you can inherit, because it was produced by somebody else's keyword list.

The test: one ad group, one set of ads, one promise

Here is the version you can apply to your own account today. Look at an ad group and read the keywords in it. Then ask: could I write one set of responsive search ad headlines that a searcher for any of these keywords would find directly relevant?

If yes, the ad group is coherent regardless of whether it holds four keywords or eighteen. If no, if you find yourself wanting headlines that pull in two different directions because two of the keywords mean different things, the ad group is doing two jobs and should be two ad groups. Not because a number was exceeded, but because one ad cannot make two different promises without weakening both.

That test also explains why the count varies so much between accounts. A narrow service with many near-synonymous phrasings can hold a lot of keywords in one coherent group. A broad catalog cannot hold three.

What happened to single keyword ad groups

The reason so many people still ask this question is that they half-remember the single keyword ad group, the structure where each keyword got its own group so the ad could match it exactly. It was a sound idea at the time, and the reason it faded is documented rather than fashionable.

Single keyword ad groups depended on exact match being literally exact. It is not. Google's documentation on keyword matching describes exact match as also covering close variants, including plurals, misspellings and closely related phrasings that the system treats as the same search. Once one keyword can trigger on a family of searches anyway, isolating it in its own ad group stops buying the precision it was built to buy. If that behavior is new to you, what each match type actually reaches covers it in full, and why your ad shows for searches you never targeted covers the consequences.

Why the quoted ranges land roughly where they do

Given all that, why do the third-party ranges cluster in single digits and low teens rather than being scattered at random? Because that is roughly the point at which the ad-copy test tends to fail on ordinary keyword lists. Somewhere past a handful of terms, most people find they are grouping keywords that no longer share one intent, and the headlines start to get vague to accommodate everything.

So the common ranges are not arbitrary, but they are also not a rule. Treat them as a signal to check rather than a limit to obey. If an ad group has twenty-five keywords and one set of headlines genuinely serves all of them, it is fine. If it has six and the headlines are already vague, it is not.

What too many keywords in one group costs you

The cost is dilution, and it is specific. When keywords in one group point at different intents, the ads have to hedge, which makes them less relevant to every one of those searches rather than less relevant to some. You also lose the ability to read the group's performance: an ad group is one of the units you look at when you are working out what is working, and a group holding four intents reports one blended number that describes none of them.

It also makes the search terms report harder to act on, because you cannot tell which of the group's several ideas let a given term through. If you are already finding that hard, the walkthrough on reading the search terms report starts from the same place.

What too few costs you, which is the small-account mistake

This side gets far less attention, and on a small budget it is the more common error. Every ad group you create divides the campaign's traffic into another bucket, and every bucket accumulates data at its own pace. Split a modest amount of traffic across a dozen tiny ad groups and no individual group ever collects enough activity to tell you anything, or to give an automated bid strategy anything to work with.

There is a real tension here, and it is worth naming rather than pretending the answer is always "tighter." Tighter groups buy relevance. Fewer, larger groups buy data density. On an account with a small budget, data density is often the scarcer resource, which argues for erring toward fewer groups than the standard advice implies, as long as the ad-copy test still passes. If you are unsure which of the two is currently costing you more, working out where the budget is actually going is the more useful place to start than a restructure.

FAQ

Does Google recommend a specific number of keywords per ad group? No. Google publishes technical limits on ad group size but no recommended count. Every specific number in circulation, including the common ranges, comes from practitioners rather than from Google.

Are single keyword ad groups completely dead? Not completely, but their original justification is gone. They were built for an era when exact match was literal, and exact match now accounts for close variants. Some advertisers still isolate individual high-value terms for control, which is a deliberate exception rather than a default structure.

Is it better to have more ad groups or more campaigns? They cost different things. Extra ad groups divide traffic and data inside a campaign. Extra campaigns divide your actual budget, because the daily budget is set at campaign level. That makes an extra ad group much cheaper than an extra campaign on a small account.

How do I know if an ad group has drifted? Read its keywords and try to write one set of headlines for all of them. If the headlines have to become vague to cover everything, the group is holding more than one intent and should be split, whatever the count happens to be.

Should I split an ad group that is performing well but has many keywords? Performing well is evidence the ad-copy test is already passing. Splitting it would divide its accumulated data across smaller groups, which on a small account is a real cost against an uncertain gain. Leave it, and spend the effort on the groups that are not working.

Checked against Google Ads' own documentation on keyword matching options, 2026-08-05. Quoted keyword-count ranges are attributed to third-party practitioner guidance, not to Google.

Similar Posts