On this page9 sections
Published October 27, 2026.
Negative keyword reviews usually happen in a panic after a spend spike. By then the wasted clicks are already paid for, the search terms report is full of noise, and someone adds 200 broad negatives that block next month's good traffic. A monthly review with a fixed window, a waste threshold, and a decision log prevents both the panic and the overcorrection.
Set the review window and the waste threshold
Use a 30-day window that ends three days before the review, so the data has settled. Define waste in currency, not just clicks. A term that spent the equivalent of two hours of a specialist's time with no conversion is a candidate; a term with 12 clicks and no conversion may simply be early.
Set the threshold with the account's economics. A high-ticket B2B account can tolerate a longer learning window than a low-margin ecommerce account. Write the threshold down so the review does not relitigate it every month.
Read the search terms report in layers
Start with terms that have spend and no conversion. Then look at terms with conversions but a poor cost per acquisition. Then inspect terms that are close variants of a converting term. The third layer is where match-type drift hides: a phrase match may be pulling in a semantically related query that the landing page cannot satisfy.
Segment by campaign and by device before deciding. A term that wastes money on mobile may be profitable on desktop. A term that fails in one geo may be the best performer in another.
Distinguish negative keywords from negative match types
A broad negative blocks more than most teams expect. Use exact and phrase negatives for surgical exclusions, and reserve broad negatives for truly irrelevant categories. Check the shared negative lists before adding a term to a campaign, or the same exclusion will be applied twice with different match types.
Audit for conflicts: a negative that blocks a keyword you are bidding on, a negative that blocks a brand term, or a negative that contradicts a competitor campaign. Google Ads can show limited diagnostics, but a manual cross-check of the keyword list is still necessary.
Protect brand and high-intent terms
Brand terms, competitor terms the team has deliberately chosen, and high-intent service terms should be on a protected list. Review the list quarterly with the business owner, not monthly with the media buyer alone. A negative added to reduce waste can accidentally block the term that carries the account.
Keep a separate note for terms that are excluded for legal or trademark reasons. Those are not performance decisions and should not be reversed because a report shows volume.
Log the decision, not just the keyword
For every change, record the term, the match type, the campaign or list, the spend and conversion evidence, the decision, the owner, and the review date. A log turns the review into an audit trail and helps a new specialist understand why a term is blocked.
Include the terms you decided not to block and why. A term that looks wasteful but is part of a longer consideration cycle may deserve one more month. Writing that down prevents the same debate next month.
Handle the "close variant" problem
Search terms reports show queries that are close variants of your keywords. A negative for one variant may not block the others. Test the negative with the platform's preview or a small campaign before applying it broadly. If the platform cannot preview, add the negative at a low-traffic campaign first and watch the impression change.
Do not assume that a negative keyword is language-specific. If the account runs multiple languages, review each language's terms separately and apply the negative to the correct campaign.
Coordinate with the landing page and the offer
Some "waste" is a message mismatch, not a targeting problem. A term that drives clicks but no conversions may be landing on a page that does not answer the query. Before blocking it, check the ad, the landing page, and the form. Blocking the term hides the mismatch and caps future demand.
Bring one landing-page hypothesis to each review. If the same mismatch appears across several terms, the page is the project, not the negative list.
Measure the effect of the change
After applying negatives, compare the next 30 days with the prior 30 on spend, clicks, conversion rate, and cost per acquisition. Segment by the campaigns that changed. A drop in spend with a stable conversion count is a good result; a drop in both may mean the negatives were too broad.
Do not judge the change in the first three days. Auction volume and learning periods create noise. Put the comparison on the next review's agenda.
Implementation checklist
- Fix the 30-day window and the waste threshold in writing.
- Review no-conversion spend, poor-CPA terms, and close variants.
- Segment by campaign, device, and geography.
- Use exact and phrase negatives before broad.
- Cross-check shared lists and keyword conflicts.
- Protect brand and legal exclusions.
- Log every decision, including terms you kept.
- Compare the next 30 days before declaring success.
