How to find wasted spend in Google Ads search terms

Most Google Ads accounts lose money in the same quiet place: searches that trigger your ads, get clicked, and never turn into a lead or a sale. Nobody notices for weeks, because no single search costs much. Added up over a month, they're often the easiest money to win back.

This guide shows how to find those searches, decide which ones are really wasted, and turn them into negative keywords, step by step.

Search terms are not keywords

A keyword is what you bid on. A search term is what someone actually typed. With phrase and broad match, one keyword can match hundreds of different searches, and some of them have nothing to do with what the business sells.

A dental clinic bidding on teeth whitening can show for:

• teeth whitening near me (good)
• teeth whitening at home kit (someone who wants to do it themselves)
• teeth whitening course (someone who wants to learn it, not buy it)
• is teeth whitening safe during pregnancy (a question, not a booking)

Each of those clicks costs the same. Only the first is likely to book.

Step 1: open the search terms report

In Google Ads, open the campaign (or the whole account), then Insights and reports → Search terms. Set the date range to the last 30 days, so there's enough data to judge.

Add these columns if they aren't there: Cost, Clicks, Conversions, Cost / conv. Then sort by Cost, highest first.

Step 2: find the terms that should have converted by now

A term with no conversions isn't wasted just because it has none yet. Five clicks with no booking proves nothing. The fair question is: given how often this campaign converts, should this term have converted by now?

Work it out from the campaign's own conversion rate:

1. Find the campaign's conversion rate for the same 30 days. Say it converts 4 in 100 clicks (4%).
2. For each term, multiply its clicks by that rate. A term with 50 clicks "should" have about 2 conversions.
3. If a term should have had two or more and has none, it's a real candidate. Fewer than that, leave it a while longer.

This keeps you from cutting good terms that simply haven't had their turn yet, and it flags the expensive ones fast.

Step 3: check whether the term fits the business

Before you exclude anything, read the term the way the client would:

• Does it match what they sell? Course, jobs, free, DIY, salary and how to are classic mismatches for a service business.
• Is it a competitor's name or a different location? Sometimes worth keeping, usually not.
• Is conversion tracking counting the right thing? If conversions are page views or phone taps instead of real enquiries, every number above is off. Check Goals → Conversions first.
• Is it new? A term that only started this week may still convert. Conversions can take days to be recorded.

If the term doesn't fit, it's wasted spend. If it fits but hasn't converted, look at the landing page and the ad before cutting it.

Step 4: add them as negative keywords

Pick the match type for each negative carefully:

• Exact match [teeth whitening course] blocks only that search. Safest.
• Phrase match "whitening course" blocks any search containing that phrase in that order.
• Broad match course blocks any search containing the word. Powerful, so use it only for words that never belong, like jobs or salary.

Add them at the campaign level when they're specific to one campaign, or to a shared negative keyword list (in the Tools menu, under the shared library) when they apply to several. A shared list for each client saves you adding the same twenty words to every campaign.

Step 5: check again next month

Search terms change with seasons, news and Google's matching. A monthly pass catches most of it; a weekly look at the top spenders catches the rest before they get expensive.

Doing this for many clients

For one account, this takes about half an hour a month. For ten clients, it's a morning, which is why it usually gets skipped.

MetricDeck does steps 1 to 3 every night for every client: it reads the search terms, judges each one against its campaign's own conversion rate, flags the ones that don't match what the business sells, and turns the ones you mark into a negative keyword list ready to paste into Google Ads. It only reads the account, so nothing changes until you add the list yourself. You can look around the demo without signing up.