adlaneSign in Practical guide
Google Ads search terms report audit: a relevance-first method
A search terms audit asks one question: do the searches behind the ad fit the offer? Judge relevance first, then cost and outcomes. A relevant term that has not converted yet deserves patience; a clearly unrelated term deserves investigation even when its current spend is small.
By Adlane · Updated
Judge relevance before cost when auditing the Google Ads search terms report
The temptation is to sort by cost and exclude whatever looks expensive. That approach removes relevant searches that simply have not had time to convert, and it leaves cheap but unrelated searches in place to grow. Start instead with relevance: can this business plausibly serve the person who typed this?
Choose a complete window long enough to contain mature conversions, and limit the review to the Search campaigns in question. Mixing campaigns with different offers makes every relevance judgement ambiguous.
Keep reported search terms separate from the keywords you configured
A keyword is something you entered into the account. A search term is something a person typed that triggered your ad. Matching rules connect the two, and the connection can be loose. Review search terms as their own list; do not assume the configured keyword tells you what the searcher wanted.
Record the keyword and match type that triggered each term where the report provides them. That pairing explains why the term appeared, and it is the first clue about whether the fix is a negative keyword, a tighter match type or a different landing page.
Classify each term as relevant, ambiguous or unrelated and record why
Give every reviewed term one of three labels and a short reason that another person could check. Note the landing page the ad sent people to, because a relevant term can still fail when the destination does not answer it. The table shows the structure.
| Classification | Test | Typical next step |
|---|---|---|
| Relevant | The offer can satisfy the search as typed | Check the ad and destination answer the question; allow time for outcomes |
| Ambiguous | The intent could be served or could be something else | Gather more clicks or inspect the landing page before deciding |
| Unrelated | The searcher wants something the business does not provide | Review a narrowly scoped negative keyword and note queries at risk |
| Eligibility mismatch | The searcher is excluded by price, region or audience | Consider messaging and targeting, not only exclusions |
Illustrative example: enterprise software and the free student download
Illustrative example: an invented business selling enterprise software sees a cluster of search terms asking for a free student download. The clicks are cheap and the click-through rate is respectable, so a cost-only sort would never flag them. The mismatch concerns eligibility and offer, not performance.
The review should consider three options rather than reaching for a universal exclusion. The ad copy could state the enterprise focus so the wrong searchers skip it. The campaign could exclude the student and free intent if the business truly has no offer for it. Or, if an education programme exists, the term could be relevant and the real problem is the landing page.
The search terms report does not contain every search
Google reports search terms that meet its privacy thresholds; low-volume queries can be omitted. Absence from the report is not proof that a query never triggered activity. Treat the reported rows as a ranked sample and compare their total clicks and cost with the campaign totals to see how much is unaccounted for.
Prioritise terms with meaningful spend or a clear mismatch. For relevant terms, inspect whether the ad and destination actually answer the searcher’s question before judging the term itself.
Run a search terms audit in Adlane
Connect the Google Ads account, then ask a bounded question with dates and a goal, such as: “Review search terms for our Search campaigns over the last complete month. Separate clearly unrelated queries from relevant queries that need more data, using enquiries as the outcome.” Adlane returns the search terms, their keyword pairing and the outcome evidence it used so you can check each classification.
Any targeting change, including a negative keyword, is a proposal that you review and approve before it is applied. New campaigns start paused. Keep the reviewed list with reasons beside each term so the decision can be revisited when the offer changes.