Find expensive nonconverting terms
Review search terms with meaningful spend and zero conversions as candidates for closer inspection.
Negative Keyword Generator
Free, no signup. Rank potential negatives by spend, zero conversions, and weak topical overlap. Review every suggestion before importing it into Ads.
Paste Ads search terms first; add GSC Queries after --- to define your organic theme. Separate data sets with a line containing ---. Files never leave your browser.
Site passport Local context for this saved site
Local data
Saved targets, named lists, and recent check summaries remain only in this browser.
Runs entirely in your browser — nothing you paste is uploaded or stored. Anonymous run-level outcome counters may be used for aggregate research; URLs, domains, IPs, and identifiers are never included, and no statistic is released below 100 runs.
Rate this tool
How to use it
- Paste an Ads search-terms export containing Search term, Cost, and Conversions.
- For safer theme filtering, add
---and paste a GSC Queries export for the relevant business area. - Select Analyze. Review suggestions from highest wasted spend downward.
- Check attribution, match behavior, close variants, and strategic value before adding any negative.
Example output Illustrative data
These fictional rows assume an organic theme around technical SEO audits.
TermSuggested matchSpendReason
free logo makerexact$86Zero conversions and little trigram overlap
jobsphrase$24Zero conversions and little trigram overlap; broad blocking risk
What you get
Wasted spend is the Ads cost in the supplied row; it is not a forecast of future savings. Exact is suggested for terms containing multiple words, while phrase is suggested for single words. The generic reason means all three model conditions passed. No finding means at least one condition failed, not that the term is proven valuable.
How it works
The engine builds a set of three-character n-grams from supplied GSC queries. Each Ads search term is converted to the same representation, and the share of its n-grams found in the theme is calculated. Terms at or below 10% overlap qualify only when cost is positive and conversions are zero. Aggregate “other” rows are excluded, and suggestions are sorted by spend.
Features
- Optional first-party organic theme instead of a generic stop-word list.
- Spend-ranked findings with conservative match-type defaults.
- Local processing and explicit Ads/GSC header validation.
Limitations
- Character overlap is lexical, not semantic; synonyms and homonyms can mislead it.
- The optional brand allowlist uses exact, contained, and typo-tolerant lexical matching; review exclusions for ambiguous brand/common-word terms.
- Conversion lag, value, assisted impact, and offline attribution are not modeled.
- Single-word phrase negatives can block many useful searches and require manual expansion checks.
Frequently asked questions
How does the negative keyword generator choose suggestions?
A search term must have positive cost, zero recorded conversions, and no more than 10 percent trigram overlap with the optional GSC query themeA free Google service that reports how a site performs in Google Search and surfaces problems with how Google crawls, indexes, and serves it. It's first-party data straight from Google — but you don't need it to appear in results..
What happens if I do not add GSC queries?
There is no theme vocabulary, so every non-aggregate Ads term with spend and zero conversions can qualify. Add GSC queries for a more selective review list.
Why are multi-word terms suggested as exact negatives?
The engine uses exact match for multi-word suggestions to reduce blocking risk. Single-word suggestions use phrase match and deserve especially careful review.
Can a zero-conversion term still be valuable?
Yes. Attribution lag, assisted conversions, low sample size, offline sales, or a strategic upper-funnel role can all make a suggestion unsafe to exclude.
Can I upload the suggestions directly to Google Ads?
Download the CSV for a review-ready negative list. The tool never changes your account or imports the list automatically.
Feature requests for Negative Keyword Generator
Upvote what you want most. New ideas can be submitted from the floating Feedback menu; requests appear here once approved, and the most-wanted rise to the top.
You won't be emailed about that request anymore.
Loading…
➕ Request a feature
New requests are reviewed before they appear here.
Where this tool helps
Common use cases
Separate off-theme traffic
Identify queries that do not fit the campaign's product, service, audience, or intent.
Prepare a negative-keyword review
Generate a candidate list for an ads manager to validate against match types, campaign structure, and business context.
Protect valuable discovery traffic
Inspect borderline terms before excluding them so a cost-saving change does not block relevant future demand.
Watch the full workflow
Negative Keyword Generator walkthrough
Read the transcript
Negative Keyword Generator
A search term with spend and no recorded conversions may be waste, delayed value, or a useful upper-funnel interaction. I’ll show you how to combine Ads terms with an optional G-S-C theme, run the local analysis, interpret spend and overlap, review match-type suggestions, use the brand allowlist, export safely, understand limitations, and make accountable next decisions.
Step 1
This local tool flags nonaggregate Ads search terms with positive cost, zero recorded conversions, and no more than ten percent trigram overlap with the optional G-S-C theme. The output is a review queue—not proof of waste and not an automated account change.
Step 2
Use it after reviewing broad or phrase match traffic, when spend accumulates outside a campaign’s intended topic, or before a negative-keyword cleanup. Adding first-party G-S-C queries makes the lexical theme more selective than cost and conversions alone.
Step 3
Paste an Ads search-terms table containing Search term, Cost, and Conversions. Add a line with three dashes, then a G-S-C Queries export for the relevant business area. Files and pasted data stay in this browser.
Step 4
This example includes two zero-conversion off-theme terms, successful S-E-O terms, and an aggregate other row. Select Analyze. The engine validates the imported headers, excludes the aggregate row, builds the G-S-C theme, and ranks qualifying candidates by supplied spend.
Step 5
The findings surface the term, suggested match type, wasted-spend input, and reason. Wasted spend is the historical cost in the row, not a forecast of future savings. No finding means at least one rule failed, not that the term is proven valuable.
Step 6
The engine converts G-S-C queries and Ads terms into three-character n-grams, then measures how much of each term appears in the theme. Ten percent or less can qualify. This is lexical comparison, so synonyms, homonyms, sparse themes, and misspellings can mislead it.
Step 7
Multiword candidates receive exact-negative suggestions to reduce blocking risk. Single-word candidates receive phrase-negative suggestions and need especially careful expansion checks because one word can block many useful searches. The suggested type is a starting point, not a command.
Step 8
Use the optional brand allowlist to exclude names, products, and common misspellings from suggestions. Matching includes exact, contained, and typo-tolerant checks. Ambiguous common-word brands still need manual review because lexical protection can be too broad or too narrow.
Step 9
Download the negative-keyword C-S-V only after inspecting the results. Review the campaign or account scope, match type, spelling, close variants, and potential conflicts with current keywords. The tool never imports or changes Google Ads for you.
Step 10
A zero can reflect conversion lag, low sample size, assisted influence, offline sales, tracking failure, or an upper-funnel role. Check the reporting window, conversion actions, values, call and C-R-M imports, and downstream behavior before classifying spend as removable.
Step 11
Character overlap is not meaning. Brand allowlisting can misclassify common words, conversion quality and assisted impact are not modeled, and a single-word phrase negative can block valuable long-tail searches. Always inspect actual query context and product relevance.
Step 12
Work from highest spend downward. Label keep, monitor, or exclude; document evidence and scope; preview affected keywords and close variants; and use a change log so decisions can be reversed. After implementation, monitor query mix, conversions, value, and unintended traffic loss.
Review the search—not just the zero.
Start with the highest-spend candidate, verify the search intent, attribution window, assisted and offline value, and exact close-variant behavior. Expand the searches a phrase negative could block, then import only reviewed rows at the right scope. Measure the change and keep a reversible decision log.
Common issues & how to fix them
- Warning Search term spent without converting Fix: Check attribution lag and business value; if the term is truly wasteful, add the reviewed negative and monitor spend.
- Warning Search term has weak theme overlap Fix: Confirm the supplied GSC theme is representative before excluding the off-theme term.
- Needs review Suggested negative may block useful traffic Fix: Do not add the single-word phrase negative until you confirm it will not block valuable searches.
- Information Multi-word term is suggested as an exact negative Fix: Review close variants, then add the multi-word term as an exact negative if its intent is irrelevant.
- Needs review Single-word term is suggested as a phrase negative Fix: Expand the single word against useful queries before adding it as a phrase negative.