A negative keyword prevents your advertisement appearing when a search contains that word. Built and maintained, it is where most advertising savings come from. Built carelessly, particularly with broad single-word exclusions, it silently suppresses searches you wanted.
Why the list matters more than the keywords
Choosing keywords tells the platform roughly what you want. Exclusions tell it what you definitely do not, and that instruction is more precise.
Without exclusions, a campaign pays for people researching how to do the job themselves, looking for employment, comparing product prices, and searching for a similarly named business in another industry. None of them were ever going to phone, and each click costs the same as one that would have.
For a small budget this is not a marginal efficiency. A campaign losing a third of its spend to irrelevant searches is a campaign operating on two thirds of its budget.
The list you can build before launching
Some exclusions are predictable in any trade and can be added before spending anything.
| Category | Examples |
|---|---|
| Doing it themselves | diy, how to, tutorial, guide, yourself, instructions, video |
| Not buying | free, cheap, cheapest, second hand, used |
| Looking for work | jobs, hiring, careers, salary, wage, apprentice, training, course, certification |
| Research and reference | meaning, definition, wikipedia, images, pictures |
| Wrong intent | rental, hire, supplies, wholesale, parts, where to buy |
That list alone prevents a meaningful share of wasted spend, and it costs nothing to add before the first click.
The list that comes from evidence
The rest cannot be predicted and comes from the search terms report, which shows what people actually typed.
The routine: open it weekly for the first month, then monthly. Read the terms that spent money and produced nothing. Add them. Repeat until the list stops growing quickly, which usually takes a couple of months.
What appears there is specific to your business in ways nobody could guess. A similarly named product, a neighbouring trade whose customers use the same words, a local place name that means something else elsewhere, or a television programme that borrowed your terminology.
Match types on the negatives themselves
The part that causes damage when it is misunderstood, because exclusions have their own matching behaviour.
A broad negative blocks any search containing that word in any order. A phrase negative blocks searches containing that sequence. An exact negative blocks only that precise search.
The practical consequence: adding a single common word as a broad negative can suppress a great deal you wanted. Excluding the word repair to avoid one bad search will remove every search containing repair, which for most trades is catastrophic.
The safe habit is to exclude phrases rather than single words wherever the single word could appear in a legitimate search, and to check what a new exclusion would have blocked over the past month before adding it.
Where to put them
- Account level shared lists for the universal ones: jobs, free, diy, and similar. Built once and applied everywhere, including future campaigns.
- Campaign level for exclusions relevant to one service but not others.
- Ad group level for separating your own groups from each other, which is how you stop one group stealing traffic intended for another.
The shared list is the one worth setting up properly, because it means every campaign you ever run starts with the accumulated learning rather than from nothing.
The over-blocking problem
The failure mode in the opposite direction, and it is harder to notice because nothing appears in a report.
A list grown enthusiastically over two years can quietly exclude terms that became relevant, or block variations of words you now want. Nothing tells you this is happening; impressions are simply lower than they should be.
Worth reviewing the list annually and asking whether each exclusion is still right. Anything added to solve a specific problem that no longer exists can come out.
Reviewing it
Weekly for the first month, monthly for the next few, then quarterly once it has stabilised. The list never quite finishes, because search behaviour changes and the platform's matching evolves.
Time spent here beats almost any other advertising activity at a small budget, because it directly reduces the cost of every enquiry rather than trying to produce more of them. How loosely the platform interprets your keywords in the first place is covered in match types and why your ads show for the wrong searches.
Frequently asked questions
What is a negative keyword?
A word or phrase that prevents your advertisement appearing when a search contains it. It is the instruction telling the platform what you definitely do not want, which is more precise than what you do.
Which negative keywords should I add before launching?
Terms indicating someone will do it themselves, is not buying, is looking for work, is doing research, or wants a different kind of transaction. That list alone prevents a meaningful share of wasted spend.
Where do the rest come from?
The search terms report, which shows what people actually typed. Read it weekly for the first month, add anything that spent money and produced nothing, and repeat until it stops growing quickly.
Can negative keywords hurt my campaign?
Yes, if added too broadly. Excluding a common single word can suppress every search containing it, which for a word like repair would remove most of your legitimate traffic.
Where should negative keywords be stored?
Universal ones in an account-level shared list so every future campaign inherits them, service-specific ones at campaign level, and separations between your own groups at ad group level.
How often should the list be reviewed?
Weekly for the first month, monthly for a few, then quarterly. It is also worth an annual check for exclusions that were added for a problem which no longer exists.
West Coast Media Solutions Inc. provides web design, web development, hosting, digital marketing, and business consulting to organisations across Canada, drawing on more than twenty-five years in the field.
Running ads without an exclusion list?
We read your search terms report, build the list from what it shows, and set it up so future campaigns inherit it automatically.
Start a Conversation