Divide Google Ads search queries into three decisions: exclude, monitor, or keep. Check search intent, conversion delay, and negative match scope with examples.
When you see a query in the Google Ads search terms report, first ask: Is this person looking for what I sell? If the intent is clearly irrelevant, the query is a candidate for a negative keyword. If the intent fits but no sale is yet visible, first check the data and conversion delay. These two situations may appear side by side in the same report and require different decisions.
First, Separate the Two Columns in the Report
Search term is the query the user typed. Keyword is the phrase you have defined in your ad account that matched the query. Your negative decision basis is the actual meaning of the query in the report, not the name of the keyword in your account. For example, a store selling only men's winter boots with the keyword “winter boots” may have had its ad triggered by the search “children's winter boots.” The campaign and ad group columns in the report show where to evaluate that query.
Google Ads shows a portion of the searches that triggered your ad in the report. Do not assume that a query not in the report had no impressions or spend. Start your review with the same campaign, date range, and conversion criteria. Clicks, cost, primary conversions, and conversion value (if any) support your decision after reading intent.
Make One of Three Decisions for Each Query
1. Exclude: Not Related to Product or Service
For example, for a store selling only ready-made boots, the query “boot repair price” is clearly irrelevant if you don't offer the service. As a first step, consider a narrow negative match covering only this query. Adding the word “repair” as a negative keyword across the entire account is a broader decision. If another campaign offers repair services, that campaign’s demand could be cut off too.
2. Monitor: Intent Fits but Data Is Insufficient
In a hypothetical case, “waterproof men's boots” got eight clicks but no purchases yet. If the product exists in the store, zero conversions alone isn't enough to exclude the query. Clicks late in the date range may not yet reflect in conversions. Check if the product is in stock, the landing page is correct, and purchase conversion is properly tracked. Then mark the query to review again after a certain date.
3. Keep: Intent and Outcome Align
Queries like “men's winter boots size 44” targeting the right product and generating sales should be kept. Before adding a broader word as a negative, check similar sales-generating queries. Excluding a word may have broader effects than just the immediate saving shown in one line.
Choose Match Type and Application Scope Together
In search campaigns, negative exact match narrows to one precise query; negative phrase match covers queries containing the specific phrase in order; negative broad match affects queries containing all chosen words in any order. Note both match type and the campaign, ad group, or list you apply the negative keyword to. Evaluate singular/plural forms and synonyms separately. Compare with sales-driving example queries side by side.
Note the implementation date. In your next review, assess both the cost saved from irrelevant queries and the conversions from relevant ones. If price, stock, or bid changes coincide with the date, don't directly attribute cost and conversion changes solely to the negative list. For regular checks, see the Google Ads negative keyword list guide.
Compare Decisions with Overall Campaign Performance
Your search query decision is based on intent and matching data in the Google Ads report. Later, you can monitor the outcome by observing spend and conversion trends at the campaign level in the AdLens analysis screen. For example, if irrelevant queries decline but campaign sales also fall, recheck the scope. Keep a simple log to track your decisions: query, decision, reason, scope, change date, and review date.
Which Rows to Check First in Weekly Reviews?
Sort the report by cost first. Mark queries generating sales separately. This method separates costly but good customer queries from budget-draining irrelevant ones. Then group rows with similar search intent. For instance, “boot repair price” and “boot sole fixing” indicate a similar service intent. If multiple terms appear, consider a shared negative phrase. Before broadening any rule, verify it doesn't overlap with sales-driving terms.
Sometimes the problem is not the query but a mismatch between the ad promise and the landing page info. For example, the query “waterproof boots” points to the right product, but if the page lacks waterproof info, first fix the page.
At the end of weekly reviews, do not measure success solely by the number of keywords added. Track whether irrelevant queries' costs decrease while conversion from relevant searches is maintained.
For example, in a hypothetical week, cost from irrelevant queries may drop from 400 TL to 100 TL, but order count from relevant queries may also fall. In this case, compare daily distributions and conversion delays between two weeks. If decrease affects only a specific product group, compare the last added rule with sales queries in that group. This narrows which changes need re-examination.
