15 most common PPC mistakes and what each one costs
The most expensive mistakes in PPC are not technical. They are decisions made on the wrong number — and in the report they look like success. A list ordered by how much money they usually cost.
- The costliest mistakes show up in the report as good numbers, not bad ones.
- The three most common: measurement that does not match, brand mixed with performance, and one return target for a whole account.
- Technical configuration mistakes usually sit at the bottom of the list by impact.
Lists of PPC mistakes usually start with the visible ones: missing negative keywords, unused extensions, the wrong match type. Those are not the expensive ones. The expensive mistakes are the ones that leave the report looking good.
This list is ordered by how much money they usually cost — not by how easily they are spotted.
Mistakes that look like success
1. Measurement that does not match orders
Nobody compared conversions in the ad system with real orders. The algorithm then dutifully produces whatever you labelled as success. Cost: every subsequent decision is inaccurate. Fix: one spreadsheet, monthly.
2. Brand mixed with performance
The account shows a ROAS of 8 because half of it comes from people searching directly for you. Cost: budget gets added against a distorted average and goes into non-brand, which may be loss-making. Fix: separate it and report in three rows. See brand vs. non-brand.
3. One return target for a whole account
Cosmetics at a 55 % margin and electronics at 9 % share a ROAS target. Cost: money flows to where things sell most easily. Fix: three margin bands in the feed. See margin by category.
4. Deciding on budget using average CPA
The account is profitable, so you add. But the added unit buys conversions dearer than the average. Cost: rising revenue with falling profit. Fix: calculate marginal CPA.
5. Judging remarketing as acquisition
It has the best numbers, so we move money into it. Cost: you pay for people who would have bought anyway. Fix: evaluate separately, and run a holdout occasionally.
Mistakes in the economics
6. The company does not know its gross margin
The whole debate about a “good cost of sale” then rests on other people’s averages. Fix: an afternoon with the accountant. No other work in PPC has a comparable return.
7. Margin calculated without returns
In fashion that changes the answer by tens of per cent. A target derived from such a margin is too loose.
8. Confusing CPA and CAC in services
An enquiry at €48 looks excellent until you find one in ten closes. Fix: CAC = CPA ÷ close rate. See metrics.
9. LTV as an excuse
“We can pay more, the customer will return.” A legitimate argument if repeat purchase is measured and counted in margin. Otherwise it is a way of approving a bigger budget.
Mistakes in running the account
10. Nobody reads the search terms report
With looser matching and AI campaigns this is now a more expensive mistake than ever. Fix: half an hour weekly. See negative keywords.
11. Dead range in the feed
Half the products with no impressions, budget concentrated on a few items. Fix: Merchant Center diagnostics. See feeds.
12. Two changes at once
You raise budget and loosen the target the same day. The result cannot be attributed. Fix: one change, a written expectation, a review date.
13. Interfering with the learning phase
After three days the change gets reverted because “it is not working”. Learning restarts. Fix: patience for two to three weeks.
14. Ads leading to the homepage
The promise in the ad appears nowhere on the page. Fix: go through the ten most expensive keywords; free and takes an hour. See landing pages.
15. No change history
Six months later nobody can establish what caused a drop or who approved what. With black boxes like Performance Max that risk is larger than before. Fix: a spreadsheet will do; anyone running several accounts needs a tool.
Mistakes that are not as costly as the talk suggests
For balance — things that get a lot of article space and little money in the account:
- Low quality score on individual keywords. Interesting as a signal, but not a metric to optimise in itself.
- Unused ad extensions. Add them, it is ten minutes’ work, but do not expect miracles.
- Low CTR. Easily raised with a discount in the headline that destroys the margin. Not a business goal.
- Choice of match type. Considerably less important than it was; conversion data and negative keywords now decide.
- Too few headline variants. Add them, but five variations of the same thing count as one to the system.
One meta-mistake
Optimising an account that measures the wrong thing. It is the most diligent way to hurt yourself.
Which is why our audits start with measurement and economics before looking at campaigns at all. The order is in the 30-point audit.
What to take away
If you take one thing from this list: check that conversions in the ad system match orders in your own system, and find out how much of the result is brand. Both take an afternoon, and both change how you read every other number.
Common questions
What is the most common mistake in PPC?
Measurement that does not match real orders, with nobody checking. It is both the most common and the most expensive, because it does not show up as bad numbers — it shows up as numbers that cannot be trusted. The algorithm then optimises towards whatever you labelled as success, even when it is not. The check is simple: once a month, compare the count and value of conversions in the ad system with real orders over the same period.
Why are my PPC costs rising even though the campaigns look fine?
Most often because decisions are being made on average cost per conversion instead of marginal. Added money goes into dearer auctions and onto less certain searches, so the last conversion added always costs more than the account average. An account can therefore be profitable overall while the added budget loses money — and in a combined report that looks like a mild dip in the average. The second common cause is brand mixed with performance, which lifts the whole average.
Which PPC mistakes matter less than people claim?
Those that get a lot of article space and little money in the account: quality score on individual keywords, unused ad extensions, low click-through rate, and nowadays even the fine choice of match type. Adding extensions takes ten minutes and should be done, but expect no miracles. Click-through rate can also be inflated with a discount in the headline that destroys the margin — it is not a business goal.