A month of paid ads, measured by nothing.
An account had been running for a month. Around £149 spent. Conversions reported: zero. The obvious conclusion was that the ads did not work, and the obvious conclusion was wrong.
Zero is a number. It is not always a result.
A zero in a conversions column reads as a verdict. Everyone who looks at it reaches the same place: the campaigns are not working, so change the campaigns. Adjust bids, rewrite ads, narrow the keywords, or stop spending.
Every one of those moves assumes the number is a measurement. It is worth about ten minutes to establish whether it is, because if it is not, each of those changes is made blind and judged by the same broken instrument that prompted it.
The distinction matters more than it sounds. "Nobody converted" and "nothing could have been recorded if they had" produce an identical report and call for opposite responses.
One says the campaign is wrong. The other says the campaign is unknown — and that every day it keeps running is a day of spend you will never be able to evaluate, whatever it produced.
Read-only first. Change nothing yet.
The audit was read-only on purpose. Altering an account before understanding it destroys the evidence of what it was doing, and the first question was not what to fix but whether the report could be believed at all.
One conversion action, and it was inactive
The account had exactly one conversion action: a generic page-view tracker. Its status was Inactive, and its lifetime conversion count was zero. It had never recorded anything, and in that state it could not.
No analytics property linked
No GA4 property was linked to the account at all. Between the two findings there was no route by which any event on the website could have reached the ads platform. The zero was not a measurement of anything.
The targeting was fine
The search-terms report was clean — the queries the ads had actually shown against were relevant, with no obvious waste to cut. Impression share was being lost to ad rank rather than to budget.
That last finding is what settles it. If the search terms had been junk, targeting would have been a plausible culprit and measurement a secondary concern. They were not. The traffic was the right traffic, arriving through an account that had no way of telling anyone what it did next.
Paused, not optimised.
Stop the spend
Campaigns paused rather than adjusted. Optimising against a broken instrument produces changes that cannot be evaluated, and the account would have accumulated a history of decisions nobody could later justify or reverse on evidence.
Rebuild the measurement from scratch
Analytics rebuilt rather than repaired. A tracking setup that has never recorded a conversion has no working state to return to, and the time spent deciding which half of it to trust exceeds the time to lay it out again properly.
Prove the events fire before spending again
Every conversion event verified as actually firing and actually arriving, checked end to end, before a single further click was paid for. A tag that loads is not a tag that reports — those are two different claims and only one of them was ever in doubt.
A rule, kept since
No paid traffic until conversion tracking is proven end to end. It is a boring rule and it costs a day at the start of an engagement. A month of unmeasurable spend costs more than a day.
Check the instrument before you trust the reading.
- +If your reports say zero conversions, establish whether that is a result or a blind spot before you touch a single bid. The two look identical and want opposite responses.
- +Open your conversion actions and read the status column. An action that is inactive, or has a lifetime count of zero, is not measuring — it is decoration.
- +Check that something is actually linked. Tracking that exists on the website but never reaches the ads platform produces a confident, detailed, entirely empty report.
- +Read the search terms before blaming the targeting. Clean queries with nothing recorded downstream is a measurement failure wearing a targeting failure's clothes.
We sell a fixed-fee Google Ads audit that starts exactly here, and one of its honest outcomes is being told to stop spending until something is fixed.
Is your zero a result, or a blind spot?
Tell us what your reports say and what you are spending, and we will tell you whether the number can be believed before anyone touches a campaign. We reply within one working day. No calls required.