Google Ads Negative Keywords: The Guide to Cutting Wasted Spend

Google Ads negative keywords are the most reliable source of savings in any paid search account, and the most consistently neglected. In the accounts we audit, somewhere between 20% and 40% of spend is going to searches that could never have converted — people looking for jobs, free versions, DIY instructions, competitors' brand names, or a completely different product that happens to share a word with yours.

Nothing about fixing this is complicated. It requires reading the search terms report on a schedule and adding exclusions. The reason it doesn't get done is that it's tedious, it makes the account's headline numbers look smaller before they look better, and Google's default settings actively push spend in the opposite direction.

What Negative Keywords Actually Do

A negative keyword tells Google not to show your ad when a search contains that term. It doesn't reduce your bid or lower your priority — it removes you from the auction entirely for that query.

This matters more now than it did five years ago. Broad match has grown steadily looser, Performance Max campaigns match on signals you can't see directly, and smart bidding will happily spend into a query pattern it believes is converting when your conversion tracking is counting something loose like a page view. Negative keywords are one of the few hard controls left.

The Three Match Types

Negative keywords have their own match types, and they behave differently from positive keyword match types in one crucial way: negatives do not match close variants. A negative for "free" will not block "frees" or a misspelling. You must add the variants yourself.

Negative match type	Syntax	Blocks	Does not block
Broad	free	Any search containing that word in any order	Misspellings, plurals, synonyms
Phrase	"free trial"	Searches containing that exact phrase in order	"trial free", misspellings
Exact	[free trial]	Only that exact search, nothing more, nothing less	"free trial software"

Broad negatives are the default choice for genuinely disqualifying terms — "job," "salary," "free," "torrent." Use phrase or exact when the word is only disqualifying in a specific combination. "Cheap" might be a hard negative for a premium law firm but a valid modifier for a discount retailer.

The most common expensive mistake is a broad negative that's too aggressive. Adding insurance as a broad negative to a car repair account will also block "does insurance cover this repair" — a high-intent query from someone with a damaged car.

Where Negatives Live: The Hierarchy

Negatives can be applied at four levels, and getting this structure right prevents the mess most accounts end up in.

1. Account-level negative lists. Terms that are disqualifying everywhere: "jobs," "careers," "salary," "free," "how to," "DIY," "torrent," "reddit" (if research-intent never converts for you), and your competitor brand names if you don't bid on them.
2. Campaign-level lists. Terms disqualifying for a particular service line but valid elsewhere. A campaign for commercial services excludes "residential"; the residential campaign excludes "commercial."
3. Ad group level. Primarily used for traffic sculpting — keeping "emergency plumber" queries out of the general plumbing ad group so they land in the emergency ad group with its own ad copy and bid.
4. Shared lists applied across campaigns. The practical way to manage account-wide negatives without duplicating them everywhere.

Build shared lists by theme rather than one giant list: a Jobs list, a Free/DIY list, a Competitors list, a Wrong-Product list. Thematic lists are far easier to audit and apply selectively than a single 900-line list nobody understands.

Not sure how much of your ad budget is going to searches that can't convert? Get a Free Growth Audit →

The Search Terms Report Workflow

This is the entire discipline. Everything else is preparation.

Frequency: weekly for accounts spending over $5,000/month, biweekly for $1,000–$5,000, monthly below that. New accounts should be checked every few days for the first month — that's when the most waste happens and the most is learned.

The process:

1. Open Search Terms, set the date range to since your last review, and sort by cost descending. Money first, not impressions.
2. Read down the list and classify each term: convert, might convert, will never convert.
3. Add every "never" to the appropriate negative list, at the appropriate level and match type.
4. For "converts well but isn't in your keyword list," add it as a keyword in the relevant ad group so you can bid on it directly.
5. For "might convert," leave it and check again next cycle with more data.

If you want a refresher on how match types and query matching interact before you start excluding things, Moz's search marketing library covers the fundamentals well.

A calibration rule: don't negate a term on one click and no conversion. Set a threshold — typically two to three times your target cost per acquisition in spend with zero conversions — before excluding. Excluding too fast on thin data is how accounts lose volume without gaining efficiency.

Note that Google no longer shows every search term. Queries below a privacy threshold are withheld, so the report is incomplete. This is another reason a well-built starter negative list matters: it catches waste you'll never see itemized.

The Starter Lists Worth Building First

If you're starting from nothing, these categories capture most of the immediate waste.

Employment intent: job, jobs, career, careers, hiring, salary, wage, resume, apply, internship, "work at." This is often the largest single waste category and it's universally disqualifying.

Free and DIY intent: free, freeware, cheap, discount, coupon, DIY, "how to," tutorial, template, sample, "do it yourself." Qualify these against your business — a discount retailer should not blanket-block "cheap."

Research intent: what is, definition, meaning, wikipedia, reddit, forum, examples, "vs" (only if comparison traffic doesn't convert for you — for many B2B advertisers it converts very well).

Wrong vertical or product: the terms that share a word with your service but mean something else entirely. A "moving company" advertiser needs negatives for "moving average," "moving on," "moving pictures." Every industry has these and they are only findable by reading your own search terms report.

Education and training: course, class, training, certification, degree, school, learn — unless you sell training.

Competitor brands: unless you're deliberately running a conquesting campaign with dedicated ad copy and landing pages, competitor names usually deliver poor click-through and worse conversion at a premium cost.

Adjacent-but-unqualified: the ones specific to your economics. A commercial-only contractor negates "residential," "home," "house," "apartment." A firm with a $10,000 minimum negates "under $500," "budget," "affordable."

Negatives in Performance Max and Broad Match

Two modern complications worth being explicit about.

Performance Max historically offered no campaign-level negative keyword control, forcing advertisers into account-level lists as the only lever. Google has since expanded negative keyword support in PMax, but the practical guidance stands: build robust account-level negative lists, because PMax will match into queries you never anticipated and the search terms visibility remains poorer than in standard search campaigns. These controls have changed repeatedly and continue to — Search Engine Journal tracks the changes closely, and you should verify current limits in your own account interface rather than trusting a guide written last year.

Broad match with smart bidding is Google's recommended default and works reasonably well — but only in an account with strong conversion tracking and a mature negative list. Running broad match on a new account with loose conversion actions and no negatives is the single fastest way to burn a budget. If you're going to use broad match, the negative list stops being maintenance and becomes the primary control surface.

Traffic Sculpting With Ad Group Negatives

Beyond blocking waste, negatives serve a second function that most accounts never use: directing queries to the right ad group.

Without ad group negatives, a query can match keywords in several ad groups, and Google picks based on its own logic rather than yours. The result is that "emergency plumber downtown" might serve the generic plumbing ad, with generic copy and a generic landing page, when you have a dedicated emergency ad group with better copy and a bid that reflects the higher value.

The technique is simple. In each ad group, add the distinguishing terms of the other ad groups as negatives:

• General ad group: negate emergency, 24 hour, same day
• Emergency ad group: negate installation, quote, cost
• Commercial ad group: negate home, residential, house

The effect is that each query lands where you intended, with matching copy and a matching bid. Click-through and conversion improve for the same traffic, and Quality Score follows because relevance improved — which is the mechanism explained in more depth in our guide to Quality Score.

Two cautions. First, sculpting is only worth doing where the ad groups genuinely warrant different copy or bids — otherwise it's complexity for its own sake. Second, with broad match and smart bidding, over-sculpting can starve the algorithm of the data it needs. Keep it to the distinctions that actually matter commercially.

Auditing an Inherited Negative List

Taking over an account with an existing negative list requires a different approach from building one, and "it already has negatives" is not the same as "it's fine." Inherited lists routinely contain terms that are quietly costing volume.

Work through it in this order:

1. Export the full list from every level — account, campaign, ad group, and each shared list. Most accounts have negatives in places nobody remembers.
2. Run the conflicts report. Google flags negatives that block your own active keywords. Fix these first; they're pure loss.
3. Look for over-broad negatives. A broad negative on a common word — cost, service, near, best — will be blocking far more than whoever added it intended. Check the volume of what each is excluding before deciding.
4. Check for negated brand variants. Someone negates a competitor name that happens to contain a word in your own brand, and you stop appearing for your own misspellings.
5. Date the assumptions. A negative added when the business only served residential customers is wrong now that they serve commercial. Business changes, lists don't.
6. Reconcile against current search terms. If a negated pattern is no longer appearing in the report anyway, it's harmless; if a heavily negated theme is now commercially relevant, unblock it deliberately.

Budget two to three hours for this on a mid-sized account. It is not unusual to find one over-broad negative suppressing 15% of available qualified volume.

Negatives at the Account Level: A Governance Note

One structural point that saves pain later: decide who is allowed to add account-level negatives, and require a reason to be recorded.

Account-level negatives are the most powerful and most dangerous exclusions in the account. A single careless entry can silently suppress an entire campaign, and because negatives don't produce error messages or performance alerts, nobody notices for weeks — they just see volume decline and start investigating bids.

The workable practice for any account with more than one person touching it:

• Account-level negatives require a documented rationale. A shared sheet with term, date, reason, and who added it.
• Shared lists are named by theme, so applying one is a deliberate decision rather than a habit.
• Campaign-level is the default. Only escalate to account level for genuinely universal exclusions.
• Review the account-level list quarterly, in full, with the rationale sheet beside it.

This sounds bureaucratic for a small account. It takes ten minutes a quarter and it prevents the single most common unexplained performance drop in paid search.

What This Is Worth

A GTA commercial equipment supplier came to us spending roughly $14,000 a month on Google Ads, with a cost per lead of $312 and a total of eleven negative keywords in the account.

Three review cycles over six weeks:

• Search terms review surfaced heavy spend on employment queries — the company's name overlapped with a well-known employer, and "jobs" variants alone accounted for about 9% of spend
• A further 16% was going to residential queries; the company sells only commercial units with a $6,000 minimum
• Around 7% went to "how to repair" and DIY-intent queries
• Competitor brand terms were running in a general campaign with generic ad copy and converting at roughly a fifth of the account average

Actions: four thematic shared negative lists built and applied, residential terms negated at account level, DIY intent negated broadly, and competitor terms moved out of the general campaign into a dedicated conquesting campaign with matched copy — rather than being deleted outright.

Outcome after eight weeks: spend held flat at $14,000, clicks down 31%, conversions up 46%, and cost per lead down from $312 to $146. No new campaigns, no new creative, no budget increase — the same money aimed at people who could actually buy.

The point of that example isn't the specific percentages. It's that the account had eleven negatives, which is functionally none, and that this is the normal state of accounts that have never been audited properly. Our paid search management work starts here for exactly that reason.

Mistakes That Cost You

• Negating too aggressively on thin data. One click, no conversion, excluded forever. Set a spend threshold.
• Conflicting negatives. A negative keyword that blocks one of your own active keywords. Google flags some of these; check the conflicts report rather than assuming.
• Forgetting variants. Negatives don't match close variants. Add plurals, common misspellings, and spacing variations ("ecommerce," "e commerce," "e-commerce").
• Adding at the wrong level. An account-level negative that kills a valid campaign is a genuinely costly error and hard to spot afterward.
• Set-and-forget. Query patterns shift constantly, especially with broad match and PMax. A list built in January is stale by April.
• Negating branded misspellings. People misspell your brand. Those are your best-converting queries.

A Sustainable Cadence

1. Once, at the start: build the four thematic shared lists and apply them.
2. Weekly or biweekly: search terms review, sorted by cost, with a spend threshold before excluding.
3. Monthly: review the conflicts report and check that no negative is blocking a converting keyword.
4. Quarterly: re-read the whole negative list. Terms negated a year ago on old assumptions sometimes deserve to come back.

This is genuinely all of it. It takes about an hour a week on a mid-sized account and it is the highest-return hour in paid search. If your account structure is also working against you, our guide on Google Ads campaign structure covers the organizational side, and Quality Score explains why tighter query matching lowers what you pay per click as a second-order effect.

Paying for clicks that were never going to become customers?

We audit your search terms report, negative list structure, and match type strategy, then quantify exactly how much of your current spend is recoverable — usually within the first week.

Get a Free Growth Audit →

---

Related reading: Google Ads Campaign Structure | Google Ads Quality Score Explained | Is Your Google Ads ROI Throwing Money Away?