How to Read the Search Terms Report in Google Ads

Jakub Pytlik | April 25, 2026 | 5 min read

What the search terms report is

When someone types "women's leather hiking boots" in Google and clicks your ad, Google records this fragment of history: which keyword in your campaign triggered, and the exact query the customer typed. The latter — the actual query text — is precisely the search terms report.

Why it matters? Because your keyword is expectation, and the search query is reality. Your campaign has "hiking boots", the customer searches "hiking boots seasonal sale 90% off". Your keyword fired, but the query that produced the click reveals the intent. Not always the one you wanted.

Where to find it in Google Ads

1. Go to Google Ads → pick a specific campaign (Search, Shopping or PMax — in Display, Video and Demand Gen this report looks different or doesn't exist)
2. In the left menu: Keywords → Search terms. Alternative for the whole-account view: Insights and reports → Search terms
3. Set the date range — I suggest the last 30 days for the full picture, 7 days for ongoing checks
4. Sort by the "Cost" column descending — to see where most of the money goes

2026 note: since 2020 Google has hidden roughly 25–30% of search queries (the low-volume ones) for privacy reasons. You only see those that crossed the threshold. So the report shows most of the traffic, but not 100%. Performance Max campaigns show even less — mainly "search themes" aggregates, not raw queries.

Query visibility limits — Google Ads Help — Search terms report. In practice about 70–75% of clicks have visible queries in Search; in PMax much less.

What the columns mean

Three things to do when looking at the report

1. Catch queries to exclude (negatives) — that's what saves the budget

The most common gain from analysing the report. Look for queries that:

• Generated > 20–30 PLN of cost without conversions
• Have an obviously different intent than your offer (e.g. "free", "torrent", "manual", "repair", "return")
• Point to outdated products / locations you don't serve
• Are semantic mistakes (e.g. "margherita pizza" when you sell pizzeria equipment)

You add these as negative keywords — either to an account-wide negatives list (e.g. "free") or to a specific campaign (e.g. "rental" for a sales-only store). From my experience, on a typical e-com account the first negatives review after 30 days of campaigns saves 15–25% of the budget — on an account with a 3,000 PLN/month budget that's a concrete 450–750 PLN per month, i.e. 5–9k PLN per year recovered from waste.

2. Catch queries to add as new keywords

The other side of the coin. Look for queries that:

• Generated conversions at a reasonable cost
• Have a clear purchase intent that you didn't have in the campaign
• Are variants of your main keyword that you don't target directly (e.g. "garage doors" + "garage door installation price")

It's worth adding these as exact match keywords in a new ad group. That way you control these queries (bid, ad, landing page) instead of leaving Google to match them at random. Important: before adding, check whether this query isn't already targeted by another, broader keyword in the same campaign. If it is — first exclude it from the broader keyword (negative at ad group level), otherwise you'll start bidding against yourself and only push costs up.

3. Check whether match type is doing its job

Look at the "Match type" column. If you see broad match generating queries far away from your keywords — that's a sign Smart Bidding is bidding too widely. From my experience, broad match combined with Maximize Conversions is one of the most common sources of wasted spend. Fix: narrow to phrase match for the main keywords + sharp negatives lists.

The most common pitfalls when reading the report

1. You look at only 7 days and draw conclusions. A query with 0 conversions in one week may earn a conversion in the second. Better: 30 days for concrete exclusion decisions, 7 days for ongoing checks.
2. You exclude everything below your conversion threshold. Some queries are upper-funnel (research) and convert on the second / third visit. Check the conversion paths in GA4 before killing a potentially sensible query.
3. You add queries as exact without checking the volume. If "garage doors automatic double-leaf" had 2 impressions in 30 days, adding it as a separate keyword does nothing.
4. You ignore match type. The same query matched broad vs exact tells a completely different story. Broad: hits widely, sometimes imprecisely. Exact: exactly as written.
5. You don't look at PMax. In Performance Max you don't see raw queries — you only have Search Themes that you suggest to Google, plus aggregated Insights. Check the Insights tab once a month: are you showing up on competitor names or on your own brand queries (brand search cannibalisation), paying more than you should. That's a common problem source on PMax accounts.

How often to look

About the sources: Report mechanics and visibility limits — Google Ads Help — Search terms report. Practical thresholds (15–25% savings after the first review) — from 11 years of running Google Ads campaigns at JP Ads. Numbers from my own practice are flagged "from my experience".