A weekly search term routine: turning waste into negative lists

Everyone who cleans the search terms report once has the same experience: three weeks later the waste is back. Broad match produces new variations, seasons change, competitors launch campaigns and users search with new words. Negative keyword management is a routine, not a project.

This article gives a practical frame for that routine: which lists, which threshold, which match type, and which chat to run every week.

Instead of keeping negatives scattered inside campaigns, build three shared lists at account level. List one is general waste: “free”, “how to”, “jobs”, “salary”, “pdf”, “used”. List two is competitor brands and marketplaces. List three is out-of-scope geography and products.

All three lists are attached when a new campaign launches. In Tecrube, “attach these three lists to all active search campaigns” is one preview and one approval; when a new term is added to a list, every campaign is protected at once.

A “non-converting term” is not always waste; the conversion window may not have closed yet. Tie the threshold to your average cost per conversion: if the average is $25, terms that spent over $25 without converting become candidates; do not judge a term that spent $10 yet.

The second criterion is intent. A high-spend term whose intent matches yours (a misspelling of your product name, say) belongs in the campaign as an exact-match keyword, not in the negative list. In chat, ask the assistant for a “negative or keyword?” suggestion with reasoning for each candidate.

For negative keywords, broad match blocks searches containing the words in any order and does not cover close variants; phrase match preserves word order; exact match blocks only that query. Single-word general patterns like “cheap” suit broad negatives; patterns like “office chair repair” suit phrase negatives.

The most frequent mistake is cutting valuable traffic with an overly broad negative. Tecrube's preview shows the match type for each negative, and you can change a row's match type before approving.

The same prompt every week: “list search terms from the last 7 days that spent over $25 without converting, group them by pattern, and suggest which shared list and match type each goes to”. The table arrives, you read the rows, remove what you disagree with, approve. In the same chat, ask “what happened to the waste ratio after last week's negatives?” to measure the effect.

If you also run Microsoft Ads, a second request carries the list over. The routine starts on Tecrube's 75 free credits; after that the Pro plan is enough for one account. Agencies repeat the same prompt for each client under the MCC; every account has its own preview and approval, visible per account in the activity log.

In search campaigns negatives are added directly; in Performance Max, campaign-level negative keyword support is limited and account-level shared lists or brand exclusions are used instead. That makes the three-layer list structure even more important here: a list attached at account level covers PMax too.

In Shopping campaigns the search terms report also shows how product titles are matching. Terms like “cheap” or “used” go to the negative list, but misspellings of a product name lead to a product title improvement rather than a negative. In Tecrube, PMax and Shopping terms are pulled with the same prompt and the suggestion arrives in the same preview format.

The goal of the routine is not a higher negative count but a lower share of non-converting spend. Once a month, ask for “non-converting spend divided by total spend” and note it. If the ratio is falling, the routine works; if it is flat, revisit the threshold or the list structure.