Your keywords are what you bid on. The search terms report shows what you actually paid for. The gap between those two things is where budget gets wasted, and where smarter accounts pull ahead.
Updated May 2026 · Written by Blake Sherman · 7 min read
Most advertisers understand keywords. The search terms report is less intuitive, but it's arguably more important to look at on a regular basis.
The term you add to your ad group and set a bid for. You control the match type. You can pause, adjust bids, or remove it at any time.
Example: "emergency plumber" at phrase match in your campaign.
The actual query a user typed into Google that matched your keyword and caused your ad to show. You don't add these; Google matches them automatically based on your match types.
Example: emergency plumber near me open now matched by your phrase keyword.
One keyword can trigger hundreds of different search terms, especially when you're using phrase match or broad match. The search terms report is the only place you can see exactly which real-world queries you spent money on.
If you're only reviewing your keyword list, you're looking at what you intended to target. The search terms report shows what actually happened.
In Google Ads: go to Keywords in the left sidebar, then click Search Terms at the top. You'll see every query that matched your keywords over the selected date range, along with clicks, impressions, cost, and conversion data.
Date range matters. A 7-day window is good for weekly hygiene. When auditing a new account or after a match type change, run it for 30-90 days to see the full picture of what's been matching.
This is what you're looking for when you review the report: a search term that isn't in your keyword list but has been spending money and converting. The question is whether to add it as a keyword, leave it matching through a broader term, or negate it. Real account data below (location redacted).
This term has been driving real revenue for a year through a broader matching keyword. At 637% ROAS it has earned its own keyword entry with a dedicated bid. Don't leave performance this strong dependent on match type spillover.
This is the best-case scenario in the search terms report. The term matched, it converted, and the numbers are strong. The action is clear: add it as an exact match keyword so you can bid on it directly, track its performance in isolation, and prevent it from getting diluted if you adjust the parent keyword later.
The flip side is also true. A term with $1,000 spent and $0 in conversion value is the same pattern in reverse. Same data, opposite call.
There are two categories of terms that need to be negated. Most advertisers handle the first category. Fewer catch the second.
Queries with no logical connection to your product or service. These are usually easy to spot. They matched because Google interpreted some keyword overlap, not because the user was anywhere close to being a customer.
These are a miss in initial campaign setup and should be caught quickly once the campaign goes live. If these are accumulating costs, your negative keyword list wasn't built before launch.
Terms that look relevant. Terms that make sense for your product. Terms that a reasonable person would assume belong in the campaign. But when you look at the data, they never convert. They just spend.
This is where most budget quietly disappears, because the natural instinct is to leave them running. They're not obviously wrong. But in practice, that budget would produce better results if it concentrated on the terms that do convert.
None of those terms look obviously wrong for a personal injury law firm. "Can I sue for personal injury" sounds like a potential client. But if that term has generated 40 clicks and zero contact form submissions across your account history, the data is telling you something.
The call is: should that budget keep funding a no-conversion term because it looks right, or should it shift to the terms that are actually producing cases? The answer is almost always to negative and redirect.
On-theme non-converters are the hardest negatives to pull the trigger on, because they require you to trust data over intuition. The term looks like it belongs. The keyword it matched makes sense. Negating it feels counterintuitive.
But "on-theme" and "converting" are not the same thing. A term can be topically aligned with your product and still represent an audience segment, intent level, or query context that doesn't produce customers for your specific offer.
Common reasons on-theme terms don't convert:
If a term has under 20-30 clicks, the conversion data isn't statistically meaningful. Don't negate based on a 10-click sample size. Set a minimum threshold of 30 clicks over a 60-day window before making a call on an on-theme non-converter. The exception: terms that are clearly research queries or career-intent terms where no amount of data will change the conclusion.
Most advertisers don't review it enough. The longer the gap, the more budget drains into terms that should have been negated three weeks ago.
| Account type | Recommended frequency | Why |
|---|---|---|
| New campaign (first 30 days) | 2x per week | Highest discovery period. Match types are casting wide, negative list is thin, spend can accumulate fast on bad terms |
| Active campaign, $50+/day spend | Weekly | At that spend level, a week of missed negatives is real money. Weekly review keeps waste contained. |
| Active campaign, <$50/day spend | Every 2 weeks | Lower stakes but still worth keeping on the calendar. Bi-weekly minimum. |
| Exact match only campaigns | Monthly | Close variants still expand exact match. Worth checking, but less urgent than phrase/broad campaigns. |
| After match type change | Within 3-5 days | Any time you expand to phrase or broad, you need to see what's matching immediately. |
One underrated habit: set a column for conversion rate alongside cost when reviewing. It puts each term's performance in context immediately rather than requiring you to mentally calculate it per row.
Build on what you've read here.
The search terms report shows the actual queries that triggered your ads and resulted in a click. It's different from your keyword list. Your keywords are what you bid on; the search terms report shows the real searches users typed. Because match types can match a wide range of queries, the search terms report is the only place you can see what you actually paid for.
For active campaigns with meaningful daily spend, weekly. For new campaigns in the first 30 days, twice a week. Monthly reviews are not frequent enough for any campaign spending real budget. A month of unreviewed terms at $100+/day adds up fast.
Two categories: clearly irrelevant terms that have no connection to your product, and on-theme terms that look relevant but consistently fail to convert. The second category is where most of the recoverable budget sits. A term can be topically close to your keywords and still represent an audience intent that doesn't convert for your specific offer.
Keywords are the terms you add to your ad groups and bid on. Search terms are the actual queries users typed that matched your keywords and triggered your ads. One keyword can match dozens of different search terms. Your search terms report shows the real-world queries; your keyword list shows what you set up to target them.
In Google Ads, go to Keywords in the left sidebar, then click Search Terms. Select the terms you want to exclude, then click "Add as negative keyword." You can add them at the ad group level or campaign level. Campaign level is right when the term is irrelevant to the whole campaign. For negatives that apply across multiple campaigns, add them to a shared negative keyword list in the Shared Library.
Google withholds search term data for queries that don't meet a minimum volume threshold, framed as a privacy measure. This means you cannot see 100% of what triggered your ads. Some spend will always appear as "(other)" in the report. The percentage of hidden data varies by account size and industry. It's a known limitation. Work with the data you can see and use tight match types and strong negative lists to minimize unknown matching.