Can you trust Google Ads conversions?
Sometimes, and you cannot tell which case you are in by looking at the number. A conversion is whatever an account was configured to count, so the figure in your report is only as meaningful as the action behind it.
Say the report from your agency shows 47 conversions this month.
The natural assumption is that those are enquiries, bookings or sales. But without knowing what is being counted, how would you know what those 47 things actually are?
They could be 47 site visits, if a page view is being recorded as a conversion.
The number itself tells you nothing until you know which action produces it. That is the question an independent check answers.
What does Google actually count as a conversion?
Whatever your setup tells it to count. Google receives a signal saying an action occurred, records it, and optimises toward it. It has no way of knowing whether that action was a customer enquiry or somebody reading your homepage.
This is the part most people assume works differently, and it is worth being precise about it. The platform is not validating that your conversions represent business outcomes. It is faithfully reporting the thing it was pointed at.
So the accuracy of your conversion data is not something Google provides. It is something your measurement setup either achieves or does not, and nobody in the usual arrangement is tasked with checking which.
What this cannot tell you Whether a given conversion action is being used as a bidding goal is a setting inside the advertising account, and no external check can see it. That distinction matters, because a miscounted action that no campaign optimises toward is a reporting problem, while the same action used as a bidding goal is a spending problem.
How can a conversion be wrong in both directions?
Overcounting happens when something that is not a customer produces a conversion signal. Undercounting happens when a real customer arrives without one being sent. Both are common, and they have opposite consequences.
Overcounting makes advertising look better than the business results justify. Reported cost per acquisition falls, campaigns look efficient, and automated bidding is taught to find more people who behave like the people who triggered the signal, who may simply have been visitors.
Undercounting is the more expensive direction, because the usual response to a campaign that looks unprofitable is to switch it off. A campaign producing real customers that go unmeasured looks exactly like a campaign producing nothing.
What this cannot tell you An external check can often see evidence of the first and evidence consistent with the second. It cannot quantify either, because that requires comparing the reports against what the business actually received, which is information only the business has.
Does a conversion number that looks reasonable mean the setup is sound?
No, and a plausible number is harder to catch than an obviously wrong one. A setup counting the wrong action produces a steady, credible-looking figure that nobody has reason to question.
The versions of this problem that get found are the ones that break loudly: conversions drop to zero, or a number triples overnight. Those get investigated within days.
The versions that persist for years are the ones that look normal. A conversion count that has always been roughly forty per month will not prompt anybody to ask what the forty things are.
What changes when the measurement is verified
This is a difference in what you know, not a promise about what your campaigns will earn. A verified number is frequently the less flattering one.
- A conversion count, with no way to tell which actions produced it.
- Reported cost per conversion, calculated against those same actions.
- A campaign ranking that assumes every counted action is comparable.
- No way to tell an underperforming campaign from a badly measured one.
- Which specific action produces each conversion, and whether your business would call it a customer.
- Which enquiry routes are measured and which are not, including phone calls.
- Whether the actions being optimised toward are the ones you would choose.
- Where the reports and your own records disagree, and by how much.
Notice that none of the right-hand column is a business result. The audit changes what you can see. What you then do about it is a campaign decision, and campaign decisions are not what this service sells.
The only way to answer this for your own account is to look.
The free check uses publicly observable evidence only, needs no account access, and tells you plainly where the evidence runs out.
How common is this, really?
Common enough to be worth checking, and this site will not put a percentage on it, because doing so honestly would require a representative sample that nobody has.
What can be said is narrower and more useful. Every measurement setup that has existed for a few years has been touched by several people with different purposes, and nothing in the normal division of labour makes anybody responsible for verifying the result. That is a structural reason to expect drift, not a statistic.
Can I check any of this myself?
Some of it, in about ten minutes, and it is worth doing before paying anybody. In Google Ads, open the conversion actions list and read what each one is named and what it counts. Two questions settle most of it: would your business call that action a customer, and is it marked as a primary action used for bidding?
Then compare a full month of reported conversions against the enquiries your business actually received. If those two numbers are far apart in either direction, you have found the thing worth investigating without needing anybody else. What is harder to check without practice is whether the ad click survives the journey to the enquiry, which is the part an external check is genuinely useful for.
What is the difference between a conversion and a customer?
A conversion is a signal your website sent. A customer is something that happened in your business. They are supposed to correspond, and the entire question is whether they actually do.
The gap between the two is where advertising money goes missing in both directions: signals sent for things that were not customers, and customers who arrived without a signal being sent.
Whose job is it to make sure conversion data is accurate?
In practice, usually nobody, which is how these situations persist for years. The advertiser assumes the agency has verified it, the agency assumes the setup it inherited was correct, and the platform reports whatever it is sent without comment.
None of those parties is being negligent. It is a genuine gap in the normal division of labour, and it is the gap this service exists to fill.