How to Read a Google Ads Search Term Report (And Why It Matters)

The short answer

The Google Ads search term report shows what users actually typed when your ad appeared. It's the closest thing PPC has to ground truth. Working from it weekly is the difference between a campaign that improves over time and one that just spends.

This post is the short version of how I read this report on every account I manage, the five things I look for, and how to turn it into a routine the team can run without me.

What the search term report actually is

Google Ads matches user searches to your bid keywords. Match types (broad, phrase, exact) define how loose or tight that match can be. The search term report shows the actual user queries that triggered your ads, alongside which of your keywords matched and the performance metrics for each.

Critical thing to understand: the search terms in the report are not the same as your bid keywords. If you're bidding on broad-match 'ultralight tent' and a user types 'how to repair a tent zipper yourself,' your ad may have shown for that query. The search term report is where you find out.

Why it's the most important report you have

Three reasons.

1. 1It exposes match-type damage. Broad and phrase match expand into queries you didn't intend. The search term report is the only place you see this happening at scale.
2. 2It surfaces high-converting queries you should break out into their own ad groups. Some of your best-performing terms are buried inside a broader match keyword. Promoting them earns higher quality scores and lower CPCs.
3. 3It's the source of every productive negative keyword. Theoretical negative-keyword lists from agencies are fine starting points. The negatives that actually save money come from queries you've already paid for that didn't convert.

How to find it

In Google Ads: Insights and reports > Search terms. (Some accounts have it under Keywords > Search terms.) Set the date range to the last 28 days minimum. Less than that and you don't have enough data to act on.

Add the columns you actually need: Search term, Keyword, Match type, Impressions, Clicks, Cost, Conversions, Conversion rate, Cost per conversion. Skip the dozens of other columns Google offers by default. Sort by Cost descending for the audit pass.

The 5 things I look for in every search term report

1. Negative keyword candidates

Queries with significant spend and zero or very few conversions. These are the easiest wins in any account. If a query has cost more than 2x your target cost-per-conversion and converted zero, it's a strong negative-keyword candidate.

• Common offenders: 'free,' 'cheap,' 'DIY,' 'how to,' 'salary,' 'jobs,' competitor brand names you don't intend to bid on.
• Add as exact-match negatives if the specific query is the problem, phrase-match negatives if a recurring word or phrase is the problem.
• Don't go nuclear on negatives the first pass. Block obvious wasters and let the next week's data show what's still leaking.

2. High-converting terms hiding inside broader keywords

Sort by Conversions descending. Look for individual search terms that converted well but are matched to a broad-match keyword that includes lots of other (less productive) variations. These are candidates to break out into their own exact-match keyword and ad group.

• If 'msr hubba hubba 2p review' converted 12 times in 28 days but is matching to a broad keyword 'ultralight tent,' add 'msr hubba hubba 2p review' as exact-match in its own ad group.
• Promoting high-performers improves quality score on those queries (relevance is tighter) and gives you ad-copy and landing-page control specific to that intent.
• Don't break out everything. Five to ten exact-match graduations per audit cycle is plenty.

3. Brand and competitor query leakage

Two patterns to watch.

• Your own brand showing up as a search term in non-brand campaigns. Means you're paying for traffic that probably would have converted organically. Add your brand as a negative across non-brand campaigns.
• Competitor brands showing up as search terms. You may or may not want to bid on these. If you do, break them out into their own campaign with appropriate ad copy. If you don't, add as negatives.

4. Match type expansion damage

Sort by spend descending and look at which match types are matching to which queries. Broad match aggressively expanded over the past two years and now matches to queries with sometimes loose semantic relationships to your bid keywords.

• If broad-match keywords are matching to queries that aren't actually relevant, narrow to phrase-match or exact-match.
• If phrase-match is matching to queries that should have been excluded, layer negatives or tighten further.
• Trust the data, not the match-type label. A keyword's match type is the upper bound on how loose Google will go. The actual queries are the truth.

5. Dead-spend patterns

Aggregate patterns of waste that aren't single-query problems. Lots of low-impression, low-conversion long-tail terms eating budget. Geographic queries you don't service. Time-of-day patterns where spend converts poorly. The search term report combined with location and time-of-day breakdowns shows you these.

• Pull the report by Location, by Hour of day, by Device. Compare conversion rate across each. If mobile is converting at 0.5% and desktop at 3%, adjust device bid modifiers.
• If queries from cities you don't service appear, add geographic negatives.
• If late-night spend isn't converting, schedule ads off during those hours.

The audit cadence

Three different reads at three different frequencies.

• Weekly: 15-minute pass. Sort by spend, sort by conversions. Add obvious negatives. Promote any clear high-converters to exact-match. Done.
• Monthly: 30-to-60-minute deeper read. Look at match-type expansion patterns, brand leakage, competitor terms. Make structural adjustments to ad groups.
• Quarterly: full search-term audit alongside the broader account audit. Re-evaluate match-type strategy, ad-group organization, and negative-keyword lists at the campaign level.

How to act on what you find

Two main actions: add negatives and promote winners.

• Negatives go on the appropriate scope. Account-level for terms you never want to match (your jobs page queries, common irrelevant terms). Campaign-level for terms that are bad for one campaign but fine for another. Ad-group-level rarely, only when the same negative shouldn't apply to a different ad group in the same campaign.
• Winners get broken out as new exact-match keywords in dedicated ad groups, with ad copy and landing pages tailored to the query. The Anatomy of a Landing Page That Converts post covers what those landing pages should look like.
• Document changes. Keep a running log of negatives added, keywords promoted, and the date. When something goes wrong (or right) in two months, you'll want the history.

Common mistakes

1. 1Adding too many negatives at once. The first audit feels productive when you flag 50 negative candidates. Add half. See what changes. Add more next week. Aggressive negative-keyword work compounds over weeks, not in one pass.
2. 2Adding negatives without considering match type. A phrase-match negative for 'job' will block 'jobs near me' and 'we are hiring jobs' and 'plumbing job application.' That's usually what you want, but confirm.
3. 3Promoting a one-conversion winner. Statistical noise. Wait for repeat performance over a meaningful window before breaking out new ad groups.
4. 4Trusting match types without auditing. Phrase-match has loosened in recent years. The search term report is the only ground truth on what's actually matching.
5. 5Auditing the search term report and never auditing the landing pages those queries hit. The whole point of finding a high-converter is sending them to a page that converts well. Run the landing page anatomy check on the destination after every promotion.

Google Ads accounts that improve over time are accounts where someone reads the search term report every week. Accounts that just keep spending are accounts where nobody does. The report is the difference.

Where this connects to landing pages

Finding a high-converting query is half the win. The other half is sending those clicks to a landing page that delivers on the promise of the query. Generic homepages and unfocused service pages waste the optimization work the search term report enables.

Pair this routine with the Landing Page Anatomy post. The two together are the bulk of what makes paid search compound: tighten the inputs, tighten the destinations.

Getting started

If you've never worked from this report before, here's the order to start.

1. 1Open Google Ads. Navigate to Insights and reports > Search terms.
2. 2Set the date range to the last 28 days. Add the columns above.
3. 3Sort by Cost descending. Identify the top 5 to 10 negative-keyword candidates (spent more than 2x target CPA, converted zero or one).
4. 4Add those negatives at the appropriate scope (account, campaign, or ad group).
5. 5Sort by Conversions descending. Identify the top 3 to 5 high-converting search terms not already broken out as exact-match keywords. Promote them.
6. 6Schedule a 15-minute weekly recurring task to repeat steps 3 to 5.
7. 7Schedule a 30-minute monthly deeper read for match-type and brand-leakage patterns.
8. 8If you want help structuring an account where this routine runs cleanly, reach out. Search-term-report routines are part of every paid search engagement I run.

Most Google Ads accounts I audit, including the outdoor and DTC brands running both Search and Shopping, have at least 15 to 25% of spend going to queries that should have been blocked. The search term report is how you find those, every week, until your account compounds instead of leaks.

Frequently asked questions

How often should I add negative keywords?

Weekly, in small batches. Adding ten negatives a week consistently outperforms adding fifty all at once and then going dormant. Negative-keyword work compounds over months. Make it a routine, not a project.

Should I use broad match at all?

Sometimes. Broad match works when you have strong conversion tracking, a tight negative-keyword list, and budget to let Google's bidding learn. It fails when any of those are missing. For most small or mid-sized accounts I work with, phrase and exact match are the safer defaults, with broad match used selectively for explicit discovery campaigns.

Why does the search term report sometimes show fewer queries than I have impressions for?

Google withholds search terms below privacy thresholds. Queries with very low search volume or queries Google considers personally identifying don't appear in the report. The aggregate clicks and cost still show, but the underlying queries don't. It's a known limitation, not a bug.

Should I delete losing keywords or just add negatives?

Depends on which is losing. If a bid keyword itself is structurally low-converting, pause or delete it. If a bid keyword is fine but is matching to specific queries that don't convert, add those queries as negatives at the right scope. The first action removes the keyword entirely. The second narrows what the keyword is allowed to match. They solve different problems.

Keep reading