Search-term review / guides

Search terms insights versus the search terms report: which one to use for negatives

By Abhishek Pal, performance marketer (apships.com). Last updated 2026-11-02.

Use the search terms report to choose negative keywords. Use search terms insights to decide which themes deserve attention. I would inspect the actual searches before choosing words to exclude. A theme can start the investigation, but it is not enough to finish the decision.

What each view shows

The search terms report shows actual searches that triggered ads and their performance. It omits some low-volume queries for privacy. Google's report documentation explains that visibility limit.

Search terms insights groups searches into categories and subcategories, with combined metrics such as clicks, conversions and conversion value. These groups include searches withheld from the term-level report, without revealing the hidden query strings. Insights is available for Search, Shopping and Performance Max. See Google's search terms insights guide.

I would choose the starting view by the question:

My questionWhere I would startWhat I need before acting
Which wording should I exclude?Search terms reportThe full searches and a business reason
Which demand theme needs attention?Search terms insightsA question to investigate across the offer and campaigns
What might my visible-row review miss?Insights alongside the reportA clear note about the limits of the evidence

A theme label is not a query record

Google says category labels are generated to represent the group and are not necessarily searches that occurred. A label containing excluded wording therefore does not, by itself, prove that the negative failed. Check the underlying terms. This distinction is covered in Google's insights FAQ.

For example, imagine a fictional lamp store sees a made-up category called "home lighting ideas". The visible searches include "brass desk lamp" and "how to make a lamp". The label describes a broad theme, but those two searches suggest different needs.

I would not add "home lighting ideas" as a negative on that evidence. I would review whether the store sells brass desk lamps and whether it wants to reach people making their own lamps. Each decision needs its own reason. A single exclusion copied from the category name skips that work.

Use hidden-term coverage without inventing hidden terms

The grouped coverage is useful because the visible export is incomplete. It still cannot tell you which unseen phrase to block. I would keep a concern about a theme separate from an exclusion supported by a visible search.

For example, suppose the fictional store sees weak conversion results in a made-up "lighting repairs" category. Some visible searches ask for repair services, which the store does not offer. Others ask for replacement lamp parts, which it sells. I would review the service searches individually and investigate the parts offer separately. I would not assign the category's whole result to the service searches.

That distinction should survive in your notes. Write which visible searches you reviewed, what you decided and what remains unknown. Do not estimate the wording or performance of hidden rows to make the review appear complete.

Prepare the right file for a negative review

This tool needs a Google Ads search-terms CSV. A category-level insights export is not a substitute. Start with the search terms export guide, then work through the CSV review process.

  1. Name the scope. Record the campaigns and dates you intend to review. Keep that note with the export.
  2. Read the full searches. Keep cost, clicks, conversions and campaign context beside the wording. Avoid reviewing an isolated word without the rows behind it.
  3. Write the business reason. Identify the product, service or intent the campaign should not pay to reach. Check that the reason holds for each proposed exclusion.
  4. Choose the negative deliberately. Review its wording, match type and destination. Check for useful searches that the same exclusion could affect.
  5. Keep unanswered questions open. If the evidence is incomplete, leave a review note. Do not turn uncertainty into a negative just to finish the list.

When insights is the better starting point

I would start with insights when choosing a broader question: whether the landing page answers the demand, whether the offer fits a theme, or whether a campaign deserves a closer term review. Those questions need a view across searches before they need a list of exclusions.

For Performance Max, use insights alongside the Performance Max search terms report, which also provides individual searches with landing-page and ad-format context. This tool sets Performance Max rows aside. Keep that review separate.

Once you have a Search or Shopping search-terms export, try the free browser check for a summary and the three most expensive negatives, with nothing uploaded. Use the full review for every suggested negative and reason, the full watchlist and export files in the $39 report. Review each suggestion; savings are estimates.

Questions

Which export should I use for a negative keyword review?
Use the search terms report, with actual search-term rows and their performance data. Keep campaign context with each row. A category summary is not enough evidence to choose the wording of a negative.
Do insights reveal searches hidden from the report?
They include those searches in grouped data, without revealing the withheld query strings. Use that coverage to investigate a theme, not to reconstruct individual hidden searches.
Does a category label prove someone searched those words?
No. Google generates labels to represent groups of searches. Read the available underlying terms before deciding what wording to exclude.
Should Performance Max reviews use insights or the report?
Use both for different questions: insights for patterns of demand, and the Performance Max search terms report for available individual searches. This tool sets Performance Max rows aside, so review them separately.

Sources

Have an export ready? The search-term review finds the wasted spend in it for free. The full report is $39. Run it on your file.

Guides