The search terms report shows the real searches people typed before they saw your ad, which is different from the keywords you added. You use it for two jobs: adding negative keywords to stop waste, and finding new keywords worth bidding on. Open it every week, sort by cost, and work down the list.
It is the most useful report in Google Ads and the one most accounts ignore. Everything else tells you how your plan performed. This one tells you what actually happened.
It also has a limitation nobody warns you about, and it changes how you should read it.
Where is the search terms report?
In the Google Ads interface, go to Campaigns, then Insights and reports, then Search terms. You can also reach it from the Keywords view by selecting keywords and choosing Search terms.
One column is off by default and you want it on. The Keyword column tells you which of your keywords matched each search. Without it you can see that you paid for a search, but not which keyword let it through, which is exactly the information you need to fix the problem. Google notes that this column has to be enabled.
Set the date range to the last sixty or ninety days. A week of data in a small account is mostly noise.
Why does the search terms report not add up to my total spend?
Because of a privacy threshold. Google only shows search terms that, in its words, “a significant number of people have used, and that resulted in your ad being shown”. Rare and unique searches are withheld.
This is the single most important thing to understand about the report. In many accounts, somewhere between a quarter and half of the spend never appears in it at all. The gap is real, it is not a bug, and there is no setting to turn it off.
What this means in practice:
- You cannot audit 100% of your spend from this report. Do not promise yourself or a client that you can.
- The hidden portion is made of long, unusual searches, which skews toward the loosest match types. Broad match hides more of its spend than exact match does.
- This is a real argument for keeping broad match in its own campaign, so that at least the campaign level totals tell you how much went to traffic you cannot inspect.
How do you use the report to cut waste?
Sort by cost, highest first, and read down until the numbers stop mattering. Ignore the terms with one impression. The money is at the top.
For each expensive term, ask one question: would I be happy to pay for this search again tomorrow? If the answer is no, add a negative keyword. If the answer is yes but the term is not converting, look at where you are sending it before you cut it.
Here is a simple triage table we use.
| What you see | Most likely cause | What to do |
|---|---|---|
| High spend, zero conversions, irrelevant term | Match type too loose | Add as a negative, check which keyword matched it |
| High spend, zero conversions, relevant term | Landing page or offer mismatch | Fix the page first, cut the term only if it stays dead |
| Low spend, good conversion rate | Under bid or too narrow | Add as its own keyword so it can get more traffic |
| Lots of “free”, “cheap”, “jobs”, “DIY” terms | Missing generic negative list | Build a shared negative list and apply it account wide |
| Competitor brand names | Broad match reaching sideways | Decide deliberately, then either block or isolate them |
What this means for you: most of the value in this report comes from the first fifteen minutes. Sort by cost, deal with the top twenty terms, repeat next week.
How do you use it to find new keywords?
Sort by conversions instead of cost, and look for terms that are converting well but are not in your keyword list. Those are searches Google found for you by accident, and they usually deserve to be found on purpose.
Adding one as its own keyword gives it a dedicated ad group, a matched ad, and a landing page chosen for it. That normally lifts both click through rate and conversion rate compared to leaving it as an accidental match.
The report is also the best source of ad copy you will find. It is your customers describing the problem in their own words. Google itself points out that the report helps you “discover new ideas for creative and landing page content”. Pull the recurring phrases straight into your headlines. Our post on responsive search ads covers how to work them in.
How often should you check it?
Weekly for a new account or one running broad match. Monthly for a stable account on exact and phrase with a mature negative list.
New campaigns need the most attention, because the first few weeks are when Google is exploring hardest and when the worst waste happens. After a few months of pruning, the well tends to run drier, and weekly checks turn up less each time. That is the sign to move to monthly.
Set a calendar reminder. This is not a task that gets done when there is spare time.
What about Performance Max and Shopping?
Performance Max does not give you a full search terms report. It gives you a search themes and categories view, which is much less detailed. You can see the general shape of the demand, not individual queries.
Shopping campaigns do give search terms, and they are worth reading, because product searches reveal a lot about how people actually describe what you sell. Feed that language back into your product titles. Our Shopping ads guide and Performance Max guide cover the reporting limits in both.
Account level negative keywords do apply to Performance Max, which is currently the main way to steer it away from traffic you do not want.
Common questions
Why do I see searches that have nothing to do with my keywords?
Broad and phrase match both work on meaning rather than words, so Google can connect concepts you would not connect yourself. Add the term as a negative and check the Keyword column to see which keyword allowed it. If one keyword keeps producing junk, it may need to move to a tighter match type. Google’s keyword matching documentation explains why broad and phrase reach further than you expect.
Should I add every irrelevant term as a negative?
No. Negative keywords block patterns, not just single searches. Adding a hundred one off negatives creates a list nobody can maintain. Look for the common word in the junk and block that instead.
Can negative keywords block too much?
Yes, and it happens more than people realize. A broad negative like “cheap” will also block “cheap alternative to [your competitor]”, which might be a customer. Check what a negative would have blocked before you add it.
Does the report show which ad was served?
Not directly in the search terms view. You can segment by other dimensions, but pairing a specific search with a specific ad is not something the report is built to do.
How far back does the data go?
You can pull historical ranges, but only for the period during which the campaign was running and served ads. Search terms below the privacy threshold were never recorded for you in the first place, so they will not appear no matter how wide you set the range.
Is there a way to see the hidden search terms?
No. The threshold is a privacy protection and there is no workaround. The best you can do is structure the account so that the unseen spend is grouped in campaigns you can judge at the campaign level. That is part of why account structure still matters.
Where to go from here
Do this today. Open the search terms report, set it to the last ninety days, sort by cost, and read the top thirty rows. Most accounts find something in that list that should have been blocked months ago.
If the list is longer than you have time for, that is usually a sign the account has been running without maintenance for a while. Our Google Ads management page explains how we handle ongoing search term work, and you can get in touch if you want us to pull the report and tell you what we see. It is a quick check and it costs you nothing.
