Short answer: Open the Google Ads search terms report, add the Cost, Conversions and Match type columns, filter for terms with spend but no conversions over the last 30 to 90 days, and sort by cost. Then group the junk into patterns (jobs, free, DIY, wrong country) and block each pattern with a shared negative keyword list instead of one-off exclusions.
What does the search terms report show, and what does it hide?
The search terms report lists the real queries people typed before they clicked your ad, not the keywords you bought. You find it in the Insights and reports menu of a Search or Shopping campaign. Each row shows the query, the keyword and match type that triggered it, plus the cost and conversion columns.
It hides two things. First, Google only lists queries that enough people searched. The rest is bundled into a row called “Other search terms”, and in broad-match-heavy accounts that row can carry a real share of the budget. You cannot exclude what you cannot see, which is why rules matter more than single exclusions.
Second, Performance Max shows far less about the queries it matches and gives you fewer negative keyword controls. That is one reason to clean up Search first; I explain the order in Performance Max vs Search, which should you run first.
How to find wasted spend in 15 minutes, step by step
- Set the date range to the last 30 to 90 days. Shorter windows hide slow patterns, longer ones mix in old settings.
- Add columns: Cost, Conversions, Cost per conversion, Conversion value, Match type and Added/Excluded. Hide impressions and click-through rate (CTR), they distract here.
- Filter for Cost greater than zero and Conversions equal to zero. Sort by cost, highest first.
- Label the top 50 rows as one of three things: irrelevant, relevant but not converting yet, or already excluded. Only the first group is wasted spend.
- Set a threshold before you exclude anything. A working rule: a zero-conversion term that has spent two to three times your target cost per acquisition (CPA) gets a decision. Say your target CPA is €30: anything above €60 to €90 with no conversion is a candidate, anything below waits.
- Flip the filter to Conversions of one or more and sort by conversions. These are the terms you protect; if they came through broad or phrase match, add them as exact keywords.
- Copy the irrelevant terms into a sheet and add a column called “pattern”. That column is where the real savings start.
Which search term patterns waste money in most accounts?
The same handful of patterns shows up in almost every Search account. Translate the example words into every language you target.
| Pattern | What you see in the report | Negative keyword rule |
|---|---|---|
| Job seekers | Your product plus “jobs”, “salary”, “career”, “hiring”, “internship” | Phrase negatives in a shared list called Jobs, applied to every campaign |
| Informational and DIY (do it yourself) | “how to”, “meaning”, “wiki”, “pdf”, “manual”, “tutorial”, “difference between” | Phrase negatives; keep them only if you sell the guide itself |
| Free and used | “free”, “used”, “second hand”, “download”, “template” | Phrase negatives, unless a free trial is your actual offer |
| Wrong buyer type | “wholesale”, “bulk”, “distributor” in a consumer shop, or “for home” in a business-to-business (B2B) account | Phrase negatives per campaign; one campaign’s junk is another’s target |
| Wrong geography or language | Cities outside your delivery area, queries in a language you cannot serve | Negatives plus a check of the location setting (presence vs presence or interest) |
| Competitor brands | Competitor names matched through broad match | Exact or phrase negatives, unless you run a deliberate competitor campaign with its own ads and budget |
| Wrong product | Your keyword plus “spare parts”, “repair”, “rental”, or a model you do not stock | Phrase negatives in a shared list called Not offered |
How do you turn one-off exclusions into negative keyword rules?
Adding “cheap red shoes free” as one exact negative kills one query. Tomorrow someone types “cheap red shoes for free” and you pay again. Rules work at the level of the word or the pattern, not the query.
Run a word-level analysis
Export the report to a spreadsheet. Split every search term into single words and build a pivot table with cost, clicks and conversions per word. Sort by cost where conversions are zero. Words like “free”, “jobs” or a city you do not serve float to the top, even when no single query containing them looked expensive on its own.
Build themed shared lists, not campaign-level piles
In the shared library, create negative keyword lists by theme: Jobs, Informational, Free and used, Wrong geography, Not offered, Competitors. Apply each list to the campaigns where it fits. When a new junk word appears, it goes into the right list once and every campaign is covered. New campaigns get the lists attached on day one.
Choose the negative match type deliberately
Negative keywords do not stretch to close variants the way positive keywords do: a negative phrase “job” does not block “jobs”. Add the singular, the plural, common misspellings and the local-language versions for each country you target. Use phrase negatives for pattern words. Use exact negatives only for one specific query. Use broad negatives sparingly: they ignore word order and block more than you intended.
Protect before you block
Before a word goes into a list, search the report for it with Conversions of one or more. If “repair” produced sales in one campaign, the negative belongs in the other campaigns’ lists, not at account level. This check keeps a cleanup from cutting revenue.
What I see in the accounts I run
When I take over an account, the negative keyword section usually looks the same: hundreds of exact-match negatives added one at a time at ad group level, most of them full queries copied from the report. The same junk pattern is still arriving, in a different word order, because exact negatives cannot catch it.
The fix is fewer negatives, not more. I rebuild them into five or six themed shared lists, mostly phrase match, with the plural and local-language variants added, and attach the lists to every Search and Shopping campaign. After that, the weekly review takes ten minutes instead of an hour, because every new find has a place to go.
In EU accounts that target several countries from one campaign, the search terms report is where the multi-country problem shows first. Queries arrive from neighbouring markets in a language the shop cannot serve, or from cities the store does not deliver to. Switching the location option from “presence or interest” to “presence” often removes more waste than any negative list.
Most common mistakes when cleaning up search terms
- Excluding after one click. Set the threshold from step five and decide on cost, not on irritation.
- Copying whole queries as exact negatives. They block one query each and pile up into a list nobody can maintain.
- Adding a broad negative at account level for a word one campaign actually sells. Check conversions per campaign first.
- Forgetting Shopping. Shopping campaigns have no keywords, so negatives are the only steering wheel you have.
- Letting auto-applied recommendations add keywords or widen match types without reading the report the following week.
- Judging leads by Google Ads conversions alone. In lead generation, check your CRM (customer database) before you exclude a term that produced enquiries.
- Confusing “not converting yet” with “irrelevant”. A relevant query with clicks and no sales is a landing page, price or tracking problem, not a keyword problem.
Start with one Search campaign and the last 90 days. Filter for cost without conversions, pick the ten most expensive irrelevant terms and write the pattern next to each one. Build your first two shared lists from those patterns today, then repeat weekly for a month.
If you would rather hand this over, it is the first thing I do in a free advertising audit, always inside your own account. You can see how I work as a PPC (pay-per-click) specialist and the full list on the services page. Email me and I’ll reply within 24 hours.
FAQ
How often should I check the Google Ads search terms report?
Weekly for the first month after launching a campaign or moving keywords to broad match, then every two to four weeks. Put it in the calendar, because the report does not alert you to anything.
Why does the search terms report not show all of my spend?
Google only lists queries that enough people searched; the rest is grouped in the “Other search terms” row. You cannot exclude what you cannot see, so themed phrase negatives protect you better than exact ones.
Do negative keywords block plurals and misspellings automatically?
No. Negative keywords do not match close variants, so add the plural, the common misspellings and the local-language versions yourself.
Should I add competitor brand names as negative keywords?
Yes in generic campaigns, unless you run a separate competitor campaign with its own ads and budget. Competitor queries matched by broad match are usually expensive clicks from people who already know what they want.
Can I add negative keywords to Performance Max?
Performance Max shows much less query detail and offers fewer negative keyword controls than Search, and the options change over time. Clean up your Search campaigns first and check the current Performance Max settings in your own account.
Is wasted spend in Google Ads always a keyword problem?
No. If relevant queries click but never convert, the leak is usually the landing page, the price or broken conversion tracking, and negatives will not fix it.