Addy
See what customers search forFind ad spend you're wastingRead paid and organic togetherPlan content from your search data
How it works
FoundersNew to search marketingProfessional servicesB2B SaaSDTC & e-commercePractitionersAgencies
PricingBlogToolsAbout
Sign InGet Started
  1. Home
  2. Blog
  3. Google Ads
  4. How to review a negative keyword list in Google Ads
Google Ads

How to review a negative keyword list in Google Ads

16 November 20268 min read
Hannah Reed

Hannah Reed

Digital strategist with over a decade in agencies and growth roles. Background in SEO and search strategy at EssenceMediaCom (WPP) and iCrossing (Hearst).

TL;DR

A negative keyword list can let through searches it was meant to stop, and stop searches the business wants. How negatives match, which entries to check first, and why adding every variant in advance isn't the answer.

In this post

  1. How a negative keyword matches
  2. Start with the negatives that reach furthest
  3. Then the ones that block too little
  4. Views on how many to add
  5. What we still don't know
  6. Sources

A negative keyword list that was built one search at a time from the search terms report, and never read as a whole, can go wrong in two directions. It can let through searches it was meant to stop, and it can stop searches the business wants.

The second is the harder one to notice. Google describes the search terms report as a list of terms "that resulted in your ad being shown", so a search that a negative blocked leaves no row in it.

How a negative keyword matches

An ordinary keyword is matched by meaning, to a degree set by its match type. Google says broad match "can include searches that don't contain the direct meaning of your keywords". A negative keyword is matched on the words typed. Google's page on them says: "Negative keywords won't match to close variants or other expansions." Casing and misspellings are handled automatically. Plurals and synonyms aren't: "you'll need to add synonyms and singular or plural versions if you want to exclude them".

Google's Ads Liaison, Ginny Marvin, has said the same in fewer words. Asked about negatives in August 2025, she replied: "Unlike positive keywords, negative keywords look only at the words themselves and not the meaning." The year before, when Google changed negatives to cover misspellings, she confirmed that "misspellings will apply to all negative keywords, regardless of match type". Both statements were reported by Barry Schwartz at Search Engine Roundtable.

There are three types, and the one applied when nobody chooses is broad. For a negative, broad means the ad "won't show if the search contains all your negative keyword terms, even if the terms are in a different order". Phrase needs the terms "in the same order". Exact needs them "in the same order, without extra words".

Two things follow from the broad rule. A one-word negative excludes every search that contains that word. A three-word negative excludes only searches that contain all three.

Negatives are worth this care because they are the one way out of a match. In the US government's case against Google the court found that "the only way to ensure that a certain query does not trigger an ad is to provide a negative keyword".

Start with the negatives that reach furthest

The order here is our judgement. It puts first the entries that can do the most damage if they are wrong.

  • Single words on broad match. Each one removes every search containing it. An invented example: a food shop adds "free" to stop searches for free samples, and also stops "sugar free" and "free delivery".
  • The account-level list. Google says it "will automatically apply to all eligible search and shopping inventory in relevant campaign types". Read literally, that covers a campaign created later as well.
  • Shared lists. A list holds "up to 5,000 negative keywords" and can be applied "to multiple campaigns at once". A word that is unwanted in one campaign may be the product in another.

Brad Geddes of Adalysis, which sells software that checks for these conflicts, wrote in a 2017 piece on negative keyword problems that they usually happen for a few reasons: a query fails in one ad group and the negative is added "at the campaign level"; "a list of negatives you always use" is applied "without examining all search queries first"; or "a shared negative list" is added "to new campaigns where that list wasn't intended". The piece is nine years old. Campaign-level negatives and shared lists work the same way today, so all three habits are still possible.

For each of these the question is the same: does any keyword, product or service in a campaign it reaches contain this word? Where the answer is yes, the negative can be narrowed to phrase or exact, or moved from the shared list to the one campaign that needed it.

Google flags some collisions itself. Its list of recommendations includes "Remove conflicting negative keywords", shown "because conflicting negative keywords are preventing some of your keywords from showing ads". It is worth acting on when it appears. The check above doesn't wait for it, and any recommendation is still worth reading before applying.

Then the ones that block too little

Google's own example is that excluding "flowers" stops an ad on "red flowers" and still allows one on "red flower". The same gap applies to synonyms. A negative such as "jobs", to take an invented example, leaves "job", "careers" and "vacancies" open, and a multi-word negative on broad match leaves open any search that drops one of its words.

The obvious fix, adding every variant in advance, runs into Google's other warning on the same page: "If you use too many negative keywords, your ads might reach fewer customers." Our rule of thumb is to add a variant once the search terms report shows it costing money, and to leave the hypothetical ones alone.

The report has a limit of its own. Google says that "some search terms that don't have enough query activity are omitted" from it. When that rule came in, in September 2020, Seer Interactive measured it across its own clients in the first days after the change and found Google Ads "hides search terms for ~28% of Paid Search budgets". That is one agency's accounts, six years ago, and Google publishes no current figure. A negative list built from the report can only cover the searches the report shows, so it will always be partial.

Views on how many to add

Nobody quoted here argues against negative keywords. The open question is how many. The evidence on it is opinion from people who work in accounts, some of whom sell software for managing negatives, a statement from Google, and one theoretical model.

For managing them closely: Geddes wrote in 2022 that as matching loosens and more accounts move to automated bidding, "negative keyword management has become one of the primary tasks of any successful PPC manager". A 2025 poll by Frederick Vallaeys, chief executive of Optmyzr, another vendor in this area, asked a narrower question: whether removing every negative from an account would make performance worse. 77% of the PPC professionals who answered thought it would. The article reporting it doesn't give the number of respondents. It shows that most practitioners want some negatives and says nothing on how many.

For restraint: Google's warning that too many negatives mean fewer customers. Google earns more when ads reach more searches, so the warning suits it, though it is also true by definition. The same Optmyzr article reports its 2024 study of more than 7,000 Performance Max campaigns, in which those using exclusions had a median cost per acquisition of $21.45, against $18.55 for those without. That is Performance Max, the exclusions included placements and topics as well as negatives, and campaigns with problems may be the ones that get exclusions added. It is a correlation from a vendor, and it cuts against that vendor's own advice.

Google's own position, in Marvin's words from the same exchange, is that a negative keyword strategy "should aim to help prevent your ads from showing on certain queries while still enabling your campaigns to achieve their goals". That sits with restraint, and it comes from the party that gains from reach.

Jyll Saskin Gales, who hosts the Inside Google Ads podcast, gave a practitioner's sense of scale in a February 2025 podcast interview: "it's not uncommon for folks to have thousands of negative keywords on search campaigns". She was discussing a limit of 100 on Performance Max and added that an advertiser needing more than that might be better off without that campaign type. The quotation is from a third-party transcript.

The argument that automation should take keyword work over has an academic version. Amaldoss, Jerath and Sayedi (Marketing Science, 2016) modelled why a search engine offers broad match with automatic bidding at all: managing keywords costs advertisers effort, that effort keeps them out of auctions, and taking it over raises the engine's profit. Their model has the engine improving the accuracy of its matching "up to the point where advertisers choose broad match", and no further. It is a theoretical model, it is ten years old and it doesn't mention negatives. What it offers is a reason to expect automation to be accurate enough to be adopted, which is a lower bar than accurate enough to need no correction.

Our reading is that both views are consistent with checking the widest negatives first. Nothing cited here tests that order.

What we still don't know

A vendor study cited in our small-business post reported that accounts using negatives converted about three times as efficiently, which is an association. How much negatives save an account, and how often one blocks a search the advertiser wanted, aren't measured by anything cited here. Google publishes no figure for the share of search terms its report leaves out. The order of review and the rule for adding variants are judgements, and a different account could reasonably choose others.

This covers Search campaigns. Google notes that "for Display and Video campaigns, all negative keywords will be considered broad matches", and other campaign types have their own rules. The same matching behaviour matters when an AI tool proposes negatives. We make software that reads search terms and negative lists, so we have an interest in people reviewing them. Everything described here can be done in the Google Ads interface, and the glossary entry has the short definition.

Sources

  • Google Ads Help: About negative keywords; About keyword matching options; About negative keyword lists; Types of recommendations; About the search terms report; Add negative keywords to campaigns. Read on 14 and 15 November 2026.
  • Brad Geddes, "How to find and fix common negative keyword problems", Adalysis, 23 February 2017, and "How to steer the Google Ads machine with negative keywords", Adalysis, 2022. A vendor's blog; practitioner commentary.
  • Optmyzr, "Negative Keywords in Google Ads", updated 30 September 2026, for the 2025 poll by Frederick Vallaeys and the 2024 Performance Max study. A vendor's article; the study is correlational.
  • Seer Interactive, "Google Ads Removes Search Terms for 28% of Paid Search Budgets", September 2020. One agency's client data.
  • Barry Schwartz, "Google Ads On If Negative Keywords Train Smart Bidding", Search Engine Roundtable, 6 August 2025, and "Google Ads Changes Ads Query Matching & Brand Controls", 27 June 2024, for the statements by Ginny Marvin, Google Ads Liaison.
  • Wilfred Amaldoss, Kinshuk Jerath and Amin Sayedi, "Keyword Management Costs and 'Broad Match' in Sponsored Search Advertising", Marketing Science 35(2), 2016.
  • Jyll Saskin Gales, interviewed on Today in Digital Marketing, "Performance Max Goes Negative", 18 February 2025. Third-party transcript.
  • United States v. Google LLC, No. 20-cv-3010 (D.D.C.), memorandum opinion of 5 August 2024, finding of fact 278.
Previous

How long to wait before judging a change in Google Ads

Related articles

Google Ads

How long to wait before judging a change in Google Ads

Google Ads reports a conversion on the day of the click, so a recent week shows all of its cost and only some of its conversions. How to measure the delay in your own account with the Days to conversion segment, and when a week is old enough to compare.

15 November 2026·4 min read
Google Ads

Should you bid on your own brand name? The evidence, and a way to check

Experiments on brand ads range from no measurable benefit to a strongly positive return, and the one study that separated the cases found competitors on the page made the difference. What to look at first, and what a pause test can and can't show.

15 November 2026·7 min read
Google Ads

How to check a Google Ads recommendation before you apply it

Each Google Ads recommendation states its own reason, in terms of traffic and conversions. None of the 88 reasons we read mentions profit. Three questions to ask before applying one, and why dismissing a recommendation raises the optimisation score as well.

15 November 2026·7 min read

Tools

  • SEO checklist
  • Technical SEO crawler
  • Search benchmarks
  • Glossary
  • All tools

Guides

  • Google Ads for small business
  • Create a Google Ads account
  • Verify Search Console
  • All guides

Topics

  • Google Ads
  • Technical SEO
  • Search strategy
  • Conversion tracking

Product

  • How it works
  • Pricing
  • About
  • Get started

Solutions

  • See what customers search for
  • Find ad spend you're wasting
  • Read paid and organic together
  • Plan content from your search data

Who it's for

  • New to search
  • Professional services
  • B2B SaaS
  • E-commerce

© 2026 Addy

·

Not affiliated with Google

PartnersContactPrivacyTerms