Negative Keywords in Google Ads: Match Types, Lists and How to Stop Wasted Spend

What are negative keywords in Google Ads?

Negative keywords are exclusion terms in Google Ads that tell a campaign which searches should never trigger your ads. You add a word or phrase, and Google stops showing your ad for queries that contain it. As a result, you stop paying for clicks from people who have no intention of buying.

The definition sounds simple. In practice, however, this is where most accounts quietly lose money. I have worked with ad accounts since 2012, and many of the accounts I take over either have no negative list at all or carry a handful of terms someone added years ago and never reviewed. Also, as broad match and automated campaign types spread, exclusions matter more than they used to.

This guide does not repeat how to read the search terms report; I cover that in my search terms report analysis article. Instead, the focus here is the exclusion layer itself: negative match types, shared lists, campaign and account level, Performance Max negatives, and how over-excluding can cut sales.

How do negative keywords stop wasted ad spend?

In Google Ads, clicks consume your budget. So an irrelevant search turns into cost the moment it triggers your ad and someone clicks. Negative keywords break that chain at the very start: the ad never enters the auction for that query, so no click and no cost follow.

The typical sources of waste look similar in almost every industry. For example, these query types rarely bring sales:

• Job seekers: "jobs", "hiring", "salary", "internship".
• Learners and researchers: "definition", "course", "pdf", "tutorial".
• People who want something free: "free", "DIY", "template download".
• Products, services or brands you do not sell.
• Locations you do not serve.

That said, the impact goes beyond money. Smart Bidding learns which searches lead to conversions, and irrelevant traffic muddies that learning. A clean negative structure gives the system better signals to work with. For that reason, I treat negatives as a data quality tool as much as a cost saving tool.

What negative match types are there?

Negative keywords also come in three match types: broad, phrase and exact. However, they behave differently from their positive counterparts. Google's help page on negative keywords explains the difference clearly. Here is a summary:

The key point: coverage on the negative side depends on the number of words. A one word negative broad term is very powerful. A three word negative broad term, in contrast, only works when all three words appear together. Therefore, short negatives usually do more work than long ones.

When you pick a type, ask one question: is this word irrelevant in every context, or only inside a specific pattern? If it is irrelevant everywhere, a one word broad negative is enough. When it only causes trouble in one pattern, choose phrase. If only a single generic query is the problem, exact is the safest option.

How does negative broad match work?

Negative broad match is the default; if you enter a term without symbols, you have chosen it. The rule: if the query contains every word in your negative, the ad does not serve. Word order does not matter, and extra words in between do not change the outcome.

For example, imagine an accounting firm that adds "internship" as a negative broad term. A query such as "accounting internship london" now stops triggering the ad. Yet "accounting intern" still gets through, because "intern" and "internship" are different words to the system. I come back to this below.

One word negative broad terms are the structure I use most. After all, the word that carries the wrong intent is usually a single word: "salary", "jobs", "pdf", "used". On the other hand, a two word negative broad term works more narrowly. "Free course" does not block "free trial lesson", because "course" is missing from the query. So choose the number of words deliberately, based on how wide you want the exclusion to reach.

When should you use negative phrase and negative exact match?

Negative phrase match applies when you put the term in quotation marks. If the words appear in the same order and next to each other, the ad does not serve, even when other words come before or after them. It fits cases where a single word is harmless on its own but irrelevant inside a certain pattern.

For example, a furniture store that sells office tables cannot exclude "table". Instead, it can exclude "table tennis" as a phrase negative. Then "foldable table tennis prices" stops triggering the ad, while "office table prices" stays open.

You write negative exact match in square brackets, and it is the narrowest exclusion. It stops the ad only when the whole query matches your term. I mostly use it in two situations:

1. When I want to close a very generic query but keep longer, high intent variations open. For instance, blocking only [accounting] while allowing "accounting firm near me".
2. For cross negatives, when I want to push a query out of one ad group and into another.

If you want to see how exact match behaves on the positive side, read my article on exact match in Google Ads; I also cover the positive side of phrase match in a separate guide.

Why don't negative keywords match close variants?

With positive keywords, Google also matches misspellings, singular and plural forms, synonyms and similar intent. Negatives do not get that expansion. In Google's own words, negative keywords don't match to close variants or other expansions.

The help page example makes this concrete. If you add "flowers" as a negative, a search for "red flowers" no longer shows your ad, but a search for "red flower" still can. In other words, singular and plural act as separate words. Also, a synonym will never trip your negative.

This rule creates extra work in every language. Excluding "job" does not exclude "jobs", "career" or "vacancy". Likewise, excluding "cheap" does nothing for "cheapest" or "low cost". You need to list the variants yourself.

On the other hand, this strictness has a benefit: your negatives never silently close queries you did not expect. There is no risk that Google treats a search as "similar" and blocks it by accident. Control sits entirely with you, and so does responsibility. That is why I think through every realistic form of a word before I add it; otherwise the list looks complete on paper but leaks in practice.

How do you handle plurals, misspellings and spelling variants?

Because negatives match literally, you should build lists from the forms people actually search, not from a root word. My method looks like this:

1. Add singular and plural forms separately: "job", "jobs".
2. Add related word forms that carry the same intent: "intern", "interns", "internship".
3. Check joined and split spellings: "e book" and "ebook", "diy" and "do it yourself".
4. Collect frequent misspellings from the search terms report and add them.

Building this list by guesswork is tiring, and it always misses something. So I start with real data instead. When you break the queries in the search terms report into individual words and see what each word costs, the variants people really use become visible. You can do that with the search terms n-gram analyzer.

Google also notes a few details about symbols. Negatives accept ampersands, accent marks and asterisks, while Google ignores periods. If you run ads in several languages, check each language's search terms separately, because spelling habits differ a lot between markets.

At which levels can you add negative keywords?

Negative keywords can work at four levels, and choosing the right level matters as much as the list itself:

• Ad group level: the negative affects only the keywords in that ad group. It suits cross negatives.
• Campaign level: the negative applies to every ad group in the campaign. It suits intent that is irrelevant to that campaign only.
• Negative keyword list: a shared list you create once and attach to several campaigns.
• Account level: a single list that covers search and shopping inventory across all eligible campaign types.

When I pick a level, I ask: which campaign is this search irrelevant for? If it is irrelevant for every campaign, such as job searches, the right place is the account level or a list attached to all campaigns. If it only fails for one product group, it belongs at campaign level.

A common mistake is adding everything to each campaign one by one. The same negative then sits in ten campaigns separately, and when you update one, the others stay outdated. Over time, that sprawl becomes a structure nobody can control.

How should you set up negative keyword lists?

A negative keyword list lets you manage the same exclusions for several campaigns in one place. According to Google's page about negative keyword lists, each list holds up to 5,000 negative keywords, and an account can have up to 20 lists. When you add a new term to a list, the change reaches every campaign that shares it at once.

Splitting lists by theme makes maintenance much easier. This is the split I use most often:

• Jobs and careers: jobs, hiring, salary, internship, resume.
• Information and learning: definition, course, pdf, tutorial, thesis.
• Free and low intent: free, torrent, crack, download.
• Competitor and unrelated brands: brand names you do not sell or do not want to target.
• Geography: cities and regions you do not serve.

The benefit is flexibility. You might not want the "definition" list on your brand campaign, but you will attach the jobs list everywhere. In this way, each campaign gets exactly the lists it needs. Also, give lists clear names; someone who opens the account in six months should see "Career searches" rather than "List 3".

When do account-level negative keywords make sense?

Account level negatives are for searches you never want to appear on anywhere in the account. Google's page about account-level negative keywords says the list works on search and shopping inventory and covers Search, Performance Max, Shopping, App, Smart and Local campaigns. The limit is 1,000 negative keywords per account.

I use this list for narrow, permanent exclusions: inappropriate terms you never want next to your brand, job search patterns, or product types you will never sell. To manage it, open the Admin area, go to Account settings, and expand the negative keywords section.

Still, dumping everything into this list is risky. An account level negative also affects campaigns you launch in the future. A word that is irrelevant today may matter in six months when you add a new product, and then you lose time wondering why a new campaign gets no impressions. For that reason, I keep the account level list short and handle the rest with themed lists.

How do negative keywords work in Performance Max?

For a long time, Performance Max was a closed box for negative keywords; you had to rely on the account level list or a support request. That has changed. According to Google's page on negative keywords in Performance Max, you can add negatives at campaign level or select an existing negative keyword list.

There is one important limit. Performance Max negatives apply to Search and Shopping inventory only. So you cannot use them to control impressions on YouTube, Display or Gmail. Google first announced a cap of 100 campaign level negatives per Performance Max campaign, then raised it to 10,000.

In practice, I follow this order. First, remember that permanent account level negatives already cover Performance Max. Next, add irrelevant themes from search term insights at campaign level. If you want to separate brand traffic, brand exclusions are a different tool and work differently from negative keywords. Setting up Performance Max in general is outside the scope of this article; my Google Shopping ads guide covers the product side basics.

Why do negatives behave differently in Display and Video campaigns?

In a Search campaign, Google compares a negative keyword with the query the user typed. Display and Video campaigns have no query; the system places the ad based on the content of a page or video. For this reason, Google states that negatives work differently in these campaigns.

According to Google, negatives here act more like topic exclusions. For instance, a set of negatives such as "women's pants" could block bidding on a page about women's jeans, even without an exact phrase match. Depending on the other targeting in the ad group, your ad may also occasionally appear on pages that contain excluded terms.

In short, do not expect Search level precision on the Display network. In these campaigns, content exclusions, placement exclusions and sensitive content settings often do more than negative keywords. If you want to understand how Google's visual campaigns work today, my Demand Gen transition article explains the current structure.

How do cross negatives route traffic to the right ad?

A cross negative excludes a query from one ad group or campaign so that another one captures it. The goal is not to cut budget; it is to send traffic to the right ad. When one query fits several keywords, the system decides which ad shows. Cross negatives put that decision back in your hands.

Consider a dental clinic with two ad groups: "dental implants" and "dental implant cost". You want people who ask about price to see the price focused ad and landing page. So you add "cost" and "price" as negatives to the general implant ad group. Then queries with a price word stay in the price ad group only.

Overdoing cross negatives backfires, though. In accounts that use Smart Bidding and broad match especially, restricting the system too much costs impressions. I use cross negatives only when the message or landing page is genuinely different, not just to make reports look tidy. To test how match types interact, the keyword match type tool makes the job easier.

Which negative keywords cut sales instead of waste?

Adding negative keywords is easy, but a wrong negative quietly costs sales. The blocked query never shows up in the report, so you only notice the loss if you look for it. These are the mistakes I see most often:

• Closing a wide concept with a one word broad negative. For example, "cheap" also excludes price sensitive buyers who are ready to purchase.
• Adding a negative that conflicts with your own keyword. The keyword stays in the account but never gets an impression.
• Blocking information words like "what is" everywhere. In some B2B services, the person in the research stage is a real prospect.
• Excluding competitor names without thinking, or never excluding them at all. That is a strategic decision, not a reflex.

To find conflicts, check Google Ads recommendations and the keyword status column regularly; a blocked keyword usually carries a flag there. Also, in every round of additions, check with conversion data whether a query ever produced a sale. I discuss this balance for B2B accounts in more detail in my B2B Google Ads search campaigns guide.

How do you start a negative list in a new account?

A new account has no search term data yet. Rather than launching with an empty list, it makes sense to prepare a starter list for your industry. That way, you prevent the most obvious waste in the first weeks.

When I build a starter list, I follow these steps:

1. I write down what the business does not sell: regions it does not serve, product types it does not offer, customer types it does not work with.
2. I add universal patterns without purchase intent: job searches, learning, free downloads.
3. I review Keyword Planner suggestions for the main terms and collect unrelated meanings. For example, "jaguar" can mean a car or an animal.
4. I keep the list limited to terms I am sure about and decide on uncertain ones after the first data comes in.

The last step matters because a starter list is a guess, and it changes as data arrives. I cover the other side, choosing the right keywords, in my article on keywords that drive sales. When you plan both lists together, the campaign starts cleaner from day one.

How does n-gram analysis feed your negative list?

Reading the search terms report line by line works for a few hundred rows. In a report with thousands of rows, however, the same word hides inside hundreds of different queries, and its total cost stays invisible. N-gram analysis solves this: it splits queries into single words and two or three word groups and adds up clicks, cost and conversions for each group.

For example, "used" may not look expensive in any single row. Yet the n-gram table immediately shows that it appears in a hundred different queries and adds up to serious spend. Also, words with high cost and zero conversions rise to the top of your negative candidate list.

My workflow is simple. I export the report, load it into the n-gram tool, and flag words with high cost and no conversions. Then I add each one with the right match type at the right level. You can do this in a few minutes with the n-gram analyzer. Still, the final decision belongs to someone who understands how the business actually sells, not to the tool.

How often should you maintain your negative keyword lists?

You cannot build a negative list once and walk away. Search behaviour changes, the business adds new products, and Google keeps updating how matching works. So the maintenance rhythm should depend on the account's spend and match types.

As a general framework, I use this rhythm:

• Frequent checks in the first weeks for new campaigns or campaigns that just moved to broad match.
• Regular checks for stable campaigns, for example weekly or every two weeks.
• A full review of all lists once a quarter to remove negatives that are no longer needed or that block sales.

Removing negatives matters as much as adding them. A word that was irrelevant once can become valuable when the business launches a new service. Also, the more crowded the lists get, the higher the risk of conflicts. For a quick view of overall account health, use the Google Ads audit tool, and plan the budget side with the Google Ads budget calculator.

How do you measure the effect of new negatives?

Every round of negatives is a small experiment, so you need to measure the result. Otherwise, you never learn which change worked. Before and after each round, I compare two periods of the same length.

These are the metrics I compare:

• Cost per conversion: it usually drops if you chose the negatives well.
• Conversion volume: it should hold steady or grow; if it falls, you may have blocked a query that sells.
• Click through rate: it often rises as irrelevant impressions disappear.
• Impression share: budget lost share can shrink because money now goes to relevant searches.

Watch out for one trap: seasonality and bid changes easily blur the result. If you add negatives and switch bid strategy in the same week, you cannot tell which change caused what. So I spread big changes over separate weeks. Also, log each round with a date and a short note; three months later, that note saves real time. Solid measurement starts with reliable conversion tracking, and my guide to enhanced conversions helps with that.

Do AI Max and broad match make negatives more important?

Yes, I think they do. Broad match and features like AI Max can take your ads far beyond the keywords you wrote. Managed well, that expansion brings new customers. If the exclusion layer is weak, however, the same expansion also opens the door to irrelevant searches.

Remember that negatives do not match close variants. On the positive side, the system builds more meaning relationships every day, while the negative side still works literally. Because of this asymmetry, accounts that move to automation need broader negative lists that they update more often.

That said, smothering automation with negatives is not the answer either. If hundreds of cross negatives shrink the system's room to explore, choosing broad match makes little sense. My balance: exclude what is clearly the wrong intent, and judge anything uncertain with data. I cover the impact on Search campaigns in my AI Max guide, and exclusions for dynamic ads in my DSA article.

How do my team and I manage negative keywords for clients?

When my team and I take over a Google Ads account, one of the first jobs is mapping the negative keywords structure: which lists exist, which campaigns they attach to, what sits at account level, and which negatives conflict with active keywords. In most accounts, this map quickly shows where the spend leaks.

Next, we review search term data with n-gram logic and check negative candidates with the client's sales team. After all, the person who talks to the leads on the phone knows best whether a term is truly irrelevant. Then we rebuild the lists by theme, attach them to campaigns, and set a maintenance rhythm based on the account's spend.

There is no shortcut here, but the payoff usually shows up quickly: more relevant clicks for the same budget. If you want us to handle the ongoing maintenance of your negatives together with the whole account, take a look at our Google Ads management service. We can measure together how much of your budget goes to irrelevant searches.