There is one report in Google Ads that tells you what you actually paid for, as opposed to what you intended to pay for. Most accounts have not opened it in months. That gap is where budget quietly leaks.
The keywords you add are a request. The search terms report is the record of what was delivered. With modern loose matching, the difference between the two is substantial & it grows every week you ignore it.
Set up the review properly
Do this weekly, same day, thirty minutes. Consistency matters more than duration, because negatives added late have already cost you the spend.
Open the search terms report & set the date range to the last seven days, plus a second view for the last ninety. The short window catches new drift. The long window reveals patterns too small to notice weekly.
Then sort by cost, descending. Not by impressions, not by clicks. You are looking for money, & sorting by cost puts the expensive mistakes at the top where they belong.
Read for four things
Every term falls into one of four categories, & each has an action.
- Converting terms not yet in your account. These are gifts. Add them as exact match keywords in the most relevant ad group so you can control the bid & write ads specifically for them.
- Spending terms with no conversions. The core of the exercise. If a term has accumulated meaningful cost with zero conversions & it is clearly not relevant, negate it. If it is relevant but not converting, look at the landing page before negating.
- Clearly irrelevant terms. Free, jobs, salary, second-hand, DIY, definitions, homework. Negate immediately & add the recurring ones to a shared list.
- Ambiguous terms. Terms where you cannot tell the intent. Leave them & check again next week with more data. Negating on insufficient evidence removes queries that would have converted.
Do not negate a term after three clicks. Wait until the spend on it is meaningful relative to your target cost per acquisition, or you will strangle the account with false negatives.
Build negative lists that scale
Adding negatives one campaign at a time creates an unmaintainable mess. Build shared lists instead, organised by purpose.
- Universal exclusions. Terms nobody in your business ever wants: jobs, careers, salary, free, torrent, complaints, scam. Applied to every campaign.
- Informational exclusions. What is, how to, meaning, definition, tutorial. Applied to commercial campaigns only, since you may deliberately target these elsewhere.
- Competitor exclusions. If you have decided not to bid on competitor names, list them once & apply everywhere.
- Service mismatch exclusions. Things you do not offer but which sit adjacent to what you do. This list is business-specific & grows steadily.
Use campaign-level negatives for cross-campaign separation, & ad group-level negatives only for fine control within a campaign. Keeping these three levels distinct prevents the situation where nobody can work out why a keyword stopped serving.
Match types for negatives, which behave differently
This trips people up constantly. Negative match types do not work like positive ones.
Negative broad match blocks any query containing all the words in any order, but does not block close variants or misspellings. Negative phrase blocks queries containing the exact sequence. Negative exact blocks only the precise term.
The practical implication is that you must add plurals, misspellings & word order variants explicitly. A negative for "free consultation" will not block "consultation free" under exact match. Check your list periodically against the search terms report to find variants slipping through.
What the report tells you beyond paid search
This is the part most teams miss entirely. The search terms report is the only place you see the actual language customers use, with commercial intent attached & conversion data alongside. No keyword research tool provides that combination.
Use it to feed other channels:
- Content planning. Questions appearing repeatedly in the report are content briefs with proven demand. Converting terms are bottom of funnel page ideas.
- Organic keyword targeting. Terms that convert well in paid are worth pursuing organically, where the traffic does not stop when the budget does.
- Service positioning. If people consistently search for your service using different words than you use, your website copy is fighting your customers' vocabulary.
- Product gaps. Repeated searches for something adjacent that you do not offer is market research you are already paying for.
We routinely use paid search query data to prioritise content marketing work, because it removes most of the guesswork from topic selection.
The limits of the report
Be aware that not every query appears. Terms below a privacy threshold are withheld, which means a portion of your spend is attributed to search terms you will never see. That share can be significant on accounts using broad match heavily.
The practical response is to watch the gap between total campaign spend & the spend visible in the search terms report. A widening gap is a reason to tighten match types rather than to keep guessing.
A thirty-minute weekly routine
- Open the last seven days, sorted by cost descending.
- Work down the top forty terms & categorise each one.
- Add new negatives to the appropriate shared list, not to individual campaigns.
- Add converting terms as exact match keywords in the right ad group.
- Note any term suggesting a content or landing page gap in a running document.
- Once a month, review the ninety-day view for slow patterns the weekly view misses.
Half an hour a week is a small cost against the proportion of spend most accounts waste on queries nobody reviewed. It is also the fastest way to develop a genuine feel for how your market searches.
If you would rather this ran as part of a managed programme, see our paid search service & pricing, or get in touch.