記事Google Ads
How to find wasted ad spend in Google Ads
Most accounts waste between a tenth and a third of their budget on clicks that were never going to convert. Here is where it hides, how to pull it in about twenty minutes, and how to decide what to do with what you find.
Ruslan9 min read
What counts as wasted ad spend?
Wasted ad spend is money paid for clicks that had no chance of converting. In a search account it is nearly always one of four things: search terms that do not match what you sell, keywords whose cost per conversion is far above what the sale is worth, campaigns still paying for a page that no longer works, and budget sitting in a campaign that stopped serving. None of it shows up as a line item. You have to go and pull it.
The figure people quote for the average account sits somewhere between 10% and 30% of budget. Treat that as a reason to look rather than as a number to repeat: what matters is what your own account is doing, and every check below is free.
1. Search terms with spend and no conversions
This is the biggest one, and the easiest to fix. Your keywords say what you want to show for. The search terms report says what people actually typed. The gap between the two is where the money goes.
In the Google Ads interface:
- Open Campaigns → Insights and reports → Search terms.
- Set the date range to the last 30 or 90 days — 90 if the account spends under $3,000 a month, or you will be reading noise.
- Add the columns Cost, Conversions and Conv. value.
- Filter to Conversions = 0, then sort by Cost, descending.
What you are looking at is every phrase that took money and gave nothing back. Add a minimum spend filter so you are not reading a thousand rows that cost eleven cents each: ten times your target cost per conversion is a reasonable floor. If your target CPA is $40, look at terms that spent more than $400 first.
2. Keywords above their cost-per-conversion ceiling
A keyword can convert and still lose money. Work out what a conversion is worth to you, then list the keywords costing more than that.
Open the Keywords view over the same period, add Cost, Conversions and Cost / conv., and sort by cost descending. Anything above your ceiling with at least 20–30 clicks behind it is a real signal. Below that, the numbers are too thin to act on.
3. Campaigns that stopped serving
Budget assigned to a campaign that is not delivering is not spent, but it is not working either — and it hides the fact that your actual reach is smaller than your plan. Look for campaigns limited by budget, disapproved ads, and ad groups with no eligible keywords. Impressions trending to zero over a week is the cheapest signal to watch.
4. Creative that has worn out
On the social side, the equivalent of a bad search term is an ad your audience has already seen too many times. The pattern is specific: frequency keeps climbing while cost per result rises with it. One without the other is not fatigue — rising frequency at a stable cost just means a small audience, and a rising cost at a flat frequency is usually an auction change.
What to do with what you find
For search terms, the fix is a negative keyword. The match type decides how much you block, and getting it wrong is how a cleanup turns into a traffic drop.
| Use | When | Blocks |
|---|---|---|
| Negative exact | One specific phrase wasted money, close variants are fine | That phrase only |
| Negative phrase | A qualifier makes the intent wrong wherever it appears | Any query containing that phrase in order |
| Negative broad | A word is always wrong for you — “free”, “jobs”, “salary” | Any query containing all those words, in any order |
Start narrow. A negative broad on a word that appears in a converting query is the one mistake in this whole process that costs you revenue rather than saving it. Put account-wide blockers in a shared negative list; keep campaign-specific ones on the campaign.
A worked example
An account spending $9,000 a month pulls 90 days of search terms. Filtered to zero conversions and more than $200 of spend, 23 terms remain, totalling $2,180. Reading them:
- Nine are job seekers — “salary”, “careers”, “jobs”. One negative broad each, at account level. $740.
- Seven are people looking for a free version. Negative broad on “free”, after checking that no converting query contains it. $610.
- Five are a competitor’s product name. Negative phrase, campaign level, because the brand campaign should still be allowed to bid there. $520.
- Two are genuinely relevant and simply have not converted yet. Left alone, flagged to look at again in 30 days. $310.
That is $1,870 a month redirected, from one report, in about twenty minutes. The two terms left alone are the important part: a cleanup that blocks everything it does not understand is not a cleanup.
How to run this with an AI assistant
Every step above is a read: you are pulling reports, not changing the account. That is what makes it a good first job to hand to an assistant, and it is why Adako charges nothing for it on any plan.
With Google Ads connected to Adako, the whole audit is one question:
Which search terms spent money in the last 90 days and never converted?
You get the terms with their spend, the total, and a proposed negative keyword for each with a match type and the reason for it. Nothing is added to the account at that point. The negatives arrive as a proposal with a field-level diff; you approve the ones you want, it executes once, and Adako reads the list back from Google so you can see what was actually stored.
The judgement calls stay yours. An assistant can tell you that a term spent $610 and converted nothing. Whether “free” is a word your business should ever block is a question about your business.
How often should you do this?
Monthly for accounts under $10,000 a month, fortnightly above that, and always after a change in match types, a new campaign, or a shift to broader targeting — those are the moments new search terms arrive in volume. If you would rather not remember, a daily rule that emails you when cost per conversion crosses a line you set does the watching for you.
The short version
- Pull search terms, filter to zero conversions, sort by cost, set a spend floor.
- Check the conversion window before you call anything waste.
- Negative exact for a phrase, negative phrase for a qualifier, negative broad only for a word that is always wrong.
- Leave relevant terms that have not converted yet alone, and look again in 30 days.
- Then check keywords above your CPA ceiling, campaigns that stopped serving, and creative where frequency and cost are climbing together.