HOW MANY KEYWORDS SHOULD I USE FOR GOOGLE ADS: MY ANSWER AFTER HUNDREDS OF CAMPAIGNS

The most common answer to how many keywords should I use for google ads is some version of "it depends," and that's true but useless. So here's the honest version after hundreds of campaigns across every budget size I've touched: fewer than you think, almost always fewer than you want, and the instinct to add "just one more" is the single most reliable way I know to make an account worse. I'm not being contrarian for its own sake. I've watched the volume approach fail so consistently that "more" has become, in my head, a warning sign rather than a strategy.

The reason this question is so hard to answer cleanly is that the person asking it is usually counting the wrong thing and measuring against the wrong ceiling. They want a number for the account. But the account total is nearly meaningless, and the number that decides whether your campaigns work lives one level down, inside the ad group. Get that unit right and the whole question reorganizes itself. Get it wrong and no amount of keyword research saves you.

The real question isn't "how many," it's "how many per intent"

When people ask how many keywords for google ads, they picture a big list living in the account and want to know how long it should be. That framing is the first mistake. The count that matters is how many keywords share a single search intent inside one ad group, because that's the number that decides whether your ad copy and landing page can honestly match what the searcher typed. Ten thousand keywords spread across two hundred tightly themed ad groups is a healthy account; two hundred keywords crammed into one ad group is a mess, even though it's a fiftieth of the size.

So the useful version of the question is: how many closely related terms can I put in one ad group before the ad stops matching all of them? Once you count intent instead of raw terms, the answer stops being a guess and becomes a structural rule. This is the same logic behind everything I do in my broader Google Ads keyword strategy: the goal isn't coverage, it's a clean line from the query to the ad to the page, and every keyword you add to a group is another chance to break that line.

It's worth saying plainly what the platform actually allows, because that ceiling is where a lot of bloated accounts get their permission. Google's own documentation on account limits puts the ceiling in the millions of targeting items per account and tens of thousands per campaign. You will never come close, and that's exactly the trap: people read the technical maximum as a suggestion. The limit that should govern you has nothing to do with that number and everything to do with your budget.

Why "more" is almost always the wrong choice

Adding keywords feels like progress because the list gets longer and it looks like you're covering more of the market. In practice you're doing the opposite of what that feeling promises. A fixed daily budget doesn't grow when your keyword count does; it just gets divided into smaller pieces. Double the keywords and you roughly halve the clicks each one can attract, which means every term now accumulates data half as fast. Do that a few times and you arrive at the state I see constantly on inherited accounts: hundreds of keywords each pulling one click a day, none of them ever reaching the volume where you could say anything confident about them.

That thin-data problem is far more damaging now than it was a few years ago, because modern campaigns lean on the algorithm to learn. Automated bidding needs enough conversions per signal to find a pattern, and when you scatter spend across a huge keyword list, no single term feeds it enough to work with. So the bloated account doesn't just waste money on irrelevant clicks; it actively starves the system you're relying on to optimize. Fewer, better-fed keywords give the algorithm a clearer diet, and it repays you with steadier performance.

There's a relevance cost on top of the data cost. Every extra keyword you force into a group loosens the connection between the search and the ad, and looser relevance shows up as lower Quality Score and higher cost per click. This is why I treat pruning as a core habit, not spring cleaning: cutting the terms that don't earn their place is often the fastest way to lift an account. If you've never audited a list for this, my process for spotting and redundant keywords in Google Ads is where I'd start, because overlap and dead weight are usually hiding in plain sight in accounts that grew by addition.

The pattern that convinced me for good is one I've now seen many times in different forms. An ad group with fifty-odd keywords, real spend flowing through it, and when you pull the report the overwhelming majority of conversions trace back to a small handful of terms. The rest weren't helping; they were diluting the data and the budget while looking like diligence. Once you've watched the 90/10 split play out on live money a few times, "add more keywords" stops sounding like optimization and starts sounding like the problem.

My actual number, and why I start below it

If you want a concrete range, here's mine: roughly 5 to 15 tightly themed keywords per ad group, and I almost always start at the low end of that, sometimes with just three or four. That's not a magic figure and I'd distrust anyone who gave you one to the decimal. It's a starting posture that reflects a simple truth: it is far easier and cheaper to add a proven keyword later than to unwind the waste from a bloated launch. Starting lean costs you a little coverage on day one; starting broad costs you real budget while you wait for enough data to justify the cuts.

The reason I lean below the range rather than at it is inference. Before launch, every keyword is a guess about intent that I can't verify until traffic arrives, so I want each of those guesses to get enough clicks to actually be tested quickly. A short list on a given budget means each term reaches readable volume in days, not months. A long list on the same budget means I'm still staring at one-click-per-keyword noise weeks later, having learned almost nothing. Understanding the types of keywords in Google Ads and how match types widen or tighten a term's reach lets me cover more search variation with fewer literal keywords, which is another reason the count can stay low without leaving gaps.

The budget tie is the part most "ideal number" advice skips, and it's the part that actually decides the answer. On a small budget, ten keywords might already be too many to feed properly; on a large one, fifty could be fine. The right count is whatever leaves each keyword getting enough clicks to produce data you can read within a couple of weeks. That's why I never quote a number without asking about spend first. "How many keywords" is really "how many can your budget afford to learn from," and once you frame it that way the number picks itself.

When I actually add a keyword

Starting lean only works if you have a disciplined way to grow, otherwise it's just an excuse to under-build. My rule is that I add a keyword when my own data tells me to, and essentially never before. The search terms report is where that data lives: after a week or two of traffic it shows the actual queries my keywords matched, and that's where new keywords earn their way in. If real searchers are typing a specific, high-intent variation I didn't include, and it's converting, I promote it into the list deliberately. That's an evidence-based addition, and it's the only kind I trust.

Everything else that tempts you to expand should be resisted. A keyword tool suggesting a lucrative-looking adjacent term is a hypothesis, not a reason to add. A competitor ranking for something isn't your data. The urge to "cover" a term because it feels related to the business is exactly the impulse that builds the bloated accounts I spend my time cleaning up. Expansion earned by your own conversion data compounds; expansion driven by fear of missing out just thins the account again. Start with the few terms you can defend, let each one get enough traffic to prove itself, and add only what your report hands you.

Put it all together and the answer to how many keywords you should use resolves into a discipline rather than a number. Count intent inside the ad group, not terms in the account. Measure "too many" against your budget, not against Google's technical ceiling. Start below your comfort level so every keyword gets fed, and add only what the search terms report proves. After hundreds of campaigns, that's the version of the answer I'd stake performance on, and it's why "more" almost never survives contact with the data.