Search Terms Report
The search terms report shows the actual queries people typed before clicking your ad, as distinct from the keywords you chose to bid on - it is the record of what you are really buying.
Why it matters
It is the only place in the account where you see what you are really buying rather than what you intended to buy. Keywords represent your guess at intent; search terms are the market telling you what it actually searched for, and the gap between the two is often larger than advertisers expect.
What people get wrong
Treating a term with no conversions as automatic waste. For a business whose leads arrive by phone, a zero-conversion search term may simply be one whose calls were never tracked, not one that failed to produce work. Check whether calls are being recorded before pausing or negating anything based on conversion count alone.
How the report actually gets populated
Every time your ad shows and someone clicks, Google logs the literal query behind that click and attributes it to the keyword that triggered the match. Because match types other than exact allow variation, one keyword can accumulate dozens or hundreds of distinct search terms underneath it, some close to the keyword's meaning and some only loosely related.
Google does not surface every search term with equal completeness. Very low-volume queries are grouped or withheld for privacy reasons, so the report is a representative sample of what happened, not a complete transcript - useful for finding patterns, less useful for accounting for every dollar spent.
What goes wrong when it is ignored
The most damaging failure is simply never reading it. An account can run for months on broad or phrase match, matching to searches that have nothing to do with the business - a plumbing company's ad showing on searches for plumbing supply retailers, or a roofing company's ad showing on searches for roofing jobs and careers - and the spend looks like ordinary account activity until someone actually opens the report and reads the queries as sentences.
The second failure is reading it but drawing the wrong conclusion from conversion counts alone, as described above - pausing terms that never had a chance to convert because the conversion action tracking that intent path is broken, not because the term itself was poor.
The symptom that should send someone to this report is a rising cost per click or falling conversion rate with no obvious campaign change. Before assuming the auction got more competitive, read what people actually typed. It is common to find the answer is a handful of irrelevant queries eating an outsized share of clicks.
How it connects to the rest of the account
The search terms report is the raw material for negative keywords: every negative added without reference to this report is a guess, and every negative added from it is a decision based on evidence. The two exist as a pair - one shows you the problem, the other lets you act on it.
It is also the only reliable way to audit how match types are actually behaving in a specific account, since the same match type setting can pull in very different query sets depending on how much conversion tracking data is steering it.
Because it is a click-level record, its usefulness is capped by how well conversions are tracked underneath it - a search terms report paired with broken call tracking will consistently misidentify good terms as bad ones.
Frequently asked
How often should the search terms report be reviewed?
Weekly is reasonable for an active account, especially early in a campaign's life or after a match type change. Established accounts with a mature negative keyword list can be checked less frequently, but a sudden shift in cost per click or conversion rate is always a reason to look immediately.
Why do some search terms not show up in the report at all?
Google withholds or aggregates very low-volume queries for user privacy, so the report is a representative view rather than a complete log of every click. This is normal and does not indicate a tracking problem; it simply means the report is best used for pattern-finding rather than exact accounting.
Should I negate every search term that has not converted?
No. A term with no conversions may simply not have had enough clicks yet, or may be feeding a conversion path - like a phone call - that is not being tracked properly. Judge relevance and intent first, and treat conversion count as one input rather than the deciding one.