Google Ads search term audit: a decision-first workflow

Start with the decision, not the junk list

A search term audit is often treated as a hunt for obvious waste. Sort by cost, mark anything that looks wrong, add negatives, move on.

That catches easy cases. It also creates two problems. Relevant terms can be excluded because they have not converted yet, and recurring issues can be missed because the reviewer only records the negative keyword, not the reason behind it.

A better audit produces a decision record:

• what needs attention;
• why it needs attention now;
• what evidence supports the call;
• what action should follow;
• where that action should apply;
• who reviewed it.

The negative keyword is one possible action. It is not the purpose of the audit.

1. Choose a review window that matches the account

Open Campaigns > Insights and reports > Search terms. Google's search terms report documentation explains the available views and notes that some low-volume queries are omitted for privacy.

Do not use the same date range blindly for every account. Check:

• the account's conversion lag;
• whether offline or qualified-lead conversions arrive later;
• recent match-type, targeting, budget, or landing-page changes;
• enough volume to judge the query without overreacting to one click;
• the date of the previous review.

If a campaign changed three days ago, review the new queries but treat performance conclusions carefully. If an account has not been reviewed for two months, use a longer window and separate old behavior from recent behavior.

2. Export the evidence you need

At minimum, keep these columns:

• search term;
• campaign and ad group;
• matched keyword and match type;
• impressions, clicks, cost, and conversions;
• conversion value, when it reflects the business outcome.

Then add decision columns outside Google Ads:

Download the CSV audit worksheet if you want a blank version with the evidence and decision fields already separated.

Field	What to record
Intent	What the searcher appears to want
Business fit	Why the query does or does not match the offer
Priority	Review now, review later, or no action
Proposed action	Keep, monitor, exclude, adjust targeting, or investigate tracking
Scope	Ad group, campaign, list, or account
Risk	What useful traffic could be blocked
Reviewer	Who approved the decision
Decision date	When the call was made

This turns a cleanup session into an auditable workflow. It also makes disagreements useful: if two reviewers make different calls, you can inspect the reason instead of comparing two unexplained lists.

3. Prioritize terms before reading them one by one

Cost is a useful starting point, but it is not a verdict. Build the first review queue from several signals:

1. High spend with a clear intent mismatch. A query for employment, training, a product you do not sell, or a location you cannot serve may deserve immediate review.
2. High spend with uncertain intent. These terms need senior judgment before any exclusion.
3. New query patterns after a campaign change. Look for repeated modifiers and themes, not isolated strings.
4. Queries with conversions but poor downstream quality. A form submission is not always a qualified lead. Check the business outcome when that data exists.
5. Repeated low-cost terms that share one irrelevant intent. A cluster can matter even when each individual query looks small.

Do not let a zero in the conversions column make the decision for you. Relevance, statistical confidence, conversion lag, and lead quality are separate questions.

4. Classify the action, not only the term

For each term or cluster, ask these questions in order.

Does the query describe something the business sells?

Use the landing page, product catalog, service area, and client exclusions. A term can look related to the keyword while describing a different product or audience.

Is the searcher's intent compatible with the campaign?

Informational intent is not automatically bad. A guide, comparison, or pricing query can be valuable in one campaign and misplaced in another. Judge it against the campaign's job.

Is the performance evidence mature enough?

Check sample size, conversion lag, tracking changes, and qualified outcomes. If the evidence is weak, mark the term for monitoring instead of forcing a binary verdict.

What is the smallest safe action?

The useful answer may be:

• keep the term;
• monitor it until more data arrives;
• add a narrow negative keyword;
• move the exclusion to a different scope;
• tighten the keyword or campaign targeting;
• fix a landing-page mismatch;
• investigate conversion tracking or lead quality.

This distinction matters. A negative keyword cannot repair a campaign whose targeting is wrong, and it cannot tell you why relevant clicks produce poor leads.

5. Choose scope and match type carefully

Google lets you add terms from the report to an ad group, campaign, or negative keyword list. Its negative keyword guidance notes that terms added directly from the report default to negative exact match unless you change the notation.

Use the narrowest scope that fixes the problem:

• Ad group: the term conflicts with one ad group's purpose.
• Campaign: the intent is wrong for the campaign but may be useful elsewhere.
• Negative keyword list: the exclusion is genuinely reusable across the campaigns to which the list is applied.
• Account level: the intent is unsuitable for the whole account and the supported campaign types.

Then choose broad, phrase, or exact based on the traffic you intend to block. The negative keyword match-type guide shows the current behavior and common over-blocking mistakes.

6. Run a safety check before applying anything

Before applying a batch, review:

• valuable queries that contain the proposed negative;
• singular, plural, and same-meaning terms that may behave differently;
• campaign and list scope;
• existing negatives and conflicts;
• the reason attached to every ambiguous decision;
• a second reviewer for high-risk exclusions.

Google warns that excessive or poorly chosen negatives can reduce reach. The current negative keyword documentation is worth checking whenever matching behavior or campaign support changes.

7. Keep a short decision history

A useful history does not need to be elaborate. Record the term or cluster, reason, action, scope, reviewer, and date. On the next audit, you can answer:

• Is this a new issue or a repeat?
• Did the exclusion solve the problem?
• Did it suppress useful traffic?
• Does the rule still fit the client's offer?
• Which accounts or campaigns now need attention first?

For agency portfolios, the next step is a cross-account queue rather than another dashboard. See how to prioritize search-term reviews across multiple Google Ads accounts.

A copyable audit checklist

Use this sequence for each review:

1. Confirm conversion lag, tracking quality, and recent campaign changes.
2. Export the search terms and performance columns.
3. Group obvious patterns before reviewing isolated queries.
4. Put high-spend mismatches and ambiguous high-cost terms first.
5. Separate relevance from performance.
6. Choose the smallest safe action and scope.
7. Review proposed negatives against valuable queries.
8. Record the reason, reviewer, and date.
9. Revisit previous decisions and check their effect.

Start with the blank CSV audit worksheet, or use the free negative keyword checker for a second pass on up to 250 search terms. The checker separates clear exclusions from questionable cases, but you should still review every suggestion before applying it. If you prefer to work through one client account live, request a free 20-minute audit call.