847 search terms. 12 converting. What that tells you about your Google Ads account.
847 search terms triggered the ads. 12 converted. The other 835 spent $2,800 without a single recorded sale. This is what Google Ads waste looks like, and how to tell it from noise.
The business owner showed me the account with confidence.
Sales were up year over year. The campaigns had been running for 18 months. The monthly reports showed graphs trending upward. Everything felt like it was working.
I pulled the search term report and filtered for the last 6 months.
847 search terms had triggered their ads. 12 had led to any conversions at all. The other 835 had collectively spent $2,800 without a single recorded conversion.
The sales increase looked much more like the business than the ads. The growth was consistent with seasonal demand and a product improvement, and very little of it traced back to paid search. I can't prove that without a holdout test, and measuring what advertising actually causes is hard even for the largest advertisers (Lewis & Rao, 2015). But nothing in the account suggested the campaigns were driving the growth. They were running alongside it.
The account in this piece is an anonymised composite of accounts I have reviewed; the numbers are illustrative and have been changed. In those accounts, the pattern is common.
Why the search term report is the most important thing you're not looking at
Google Ads runs on intent. Someone types something into Google. Your ad is eligible to show. They click.
The keyword you're bidding on is what you told Google to target. The search term is what the person actually typed.
These are not the same thing.
A single phrase match or broad match keyword can trigger hundreds of different search terms - many of which have nothing to do with what you sell. Google gets paid per click regardless of whether that click converts. Your job is to ensure you're paying only for searches with a realistic chance of becoming a customer.
Most accounts don't do this consistently. Campaigns are set up, the ads run, and the search term report gets skipped in favour of the higher-level metrics that look more reassuring.
What was actually in that account
Here's the breakdown from the account described above.
Twelve of 847 search terms produced every recorded conversion
Each square is one search term the ads showed for, over six months.
| search terms | Spend | |
|---|---|---|
| Converting | 12 | $1,400 |
| Relevant-looking, clicked, no conversion | 315 | $2,100 |
| Clearly irrelevant | 127 | $700 |
| Impressions only, no clicks | 393 | $0 |
12 terms, every recorded conversion.
The 12 search terms doing the actual work accounted for $1,400 of spend. The other $2,800 went to terms that never converted: $2,100 to searches that looked relevant, plus $700 to searches that were clearly off-target from the start.
That's $2,800 in spend with no recorded conversion.
Over 6 months.
Not all of it is waste, and this is where most search term reviews go wrong. The 315 relevant-looking terms averaged under $7 each, a handful of clicks apiece. A term with five clicks and no sale is entirely consistent with a normal conversion rate. "No conversions in six months" tells you very little on its own.
The rule I use is a spend threshold: a term that has spent two to three times your target cost per conversion without converting has had a fair chance. Below that, it is still noise. In this account the target was around $35, and 14 of the 315 terms had crossed $70 with nothing to show for it. Together they had spent $1,300. The other 301 had spent $800 between them, and they stay on watch.
Read the rest of this piece
Subscribe to The Marketing Economist: marketing and paid media through an economics lens. The full article opens right here, along with every other one on this site. New issues arrive about once a month.
Why this happens
There's a structural reason this problem is so common.
Google's default campaign settings are designed to maximise reach. Broad match keywords, Smart Campaigns, Performance Max - all of these are built to expand who sees your ads. None of them is built to hold spend back to the terms that have already converted. The platform's incentive is more clicks. Your incentive is fewer, better clicks.
Those objectives only align when someone is actively managing the gap between them.
Without regular search term reviews and systematic negative keyword management, the gap grows. Google's definition of "relevant" and your customers' actual purchase intent drift further apart with every passing month.
The algorithm isn't doing anything wrong. It's doing exactly what it's configured to do. The configuration just hasn't been updated to reflect what's actually converting.
The warning signs you can check right now
The search term waste problem doesn't announce itself in top-level metrics. Impressions look reasonable. Click volume seems healthy. The account appears active.
The signals show up one level deeper:
- High click volume, low conversion rate: clicks rising while the paid search conversion rate drifts down
- Search terms past the spend threshold with zero conversions: terms that have each spent two to three times your target cost per conversion and produced nothing
- A negative keyword list that hasn't been updated in 3+ months: searches you should have excluded months ago are still spending
- Brand terms appearing in non-brand campaigns: competitors and branded searches running in generic campaigns, inflating costs
- Cost per conversion trending upward while lead volume stays flat: you're spending more per outcome without getting more outcomes
Any one of these usually points to recoverable budget. All of them together, which I see often in accounts that have been running on autopilot, means the account is likely spending more than it needs to for the results it's producing.
What happens when you fix it
In the account above: adding the 127 clearly irrelevant terms and the 14 terms past the threshold as negative keywords (search terms can't be paused; you exclude them) freed $2,000 of the $2,800. That $2,000 went back toward the 12 search terms that were actually working, on top of the $1,400 they already had.
Same budget. More spend on the searches that convert.
What I saw afterwards was simple: more of the same budget reached the terms with a track record of converting. The usual story is that this also gives the bidding algorithm cleaner conversion data and lifts Quality Scores. That is plausible, but I didn't observe it directly, and Quality Score is a keyword-level diagnostic, so I don't treat it as the mechanism.
The same budget, run two ways
The managed account costs the same to run. It can produce different outcomes from the same budget.
In short
- The keyword you bid on and the search term someone types are not the same thing.
- In the accounts I have reviewed that run without regular search term reviews, a large share of spend often goes to searches that never convert.
- The search term report is the most important thing most businesses aren't looking at.
- Cut on a spend threshold: a term needs to spend two to three times your target cost per conversion before its silence means anything.
- Fixing waste means adding negatives, which concentrates budget on proven terms.
- If your account has been running 6+ months without a search term review, the waste is already there.
References
- Lewis, R. A., & Rao, J. M. (2015). The unfavorable economics of measuring the returns to advertising. Quarterly Journal of Economics, 130(4), 1941–1973.
If something here was useful, would love to know!
adela@dafe.ro