If you have ever read an agency report, felt quietly pleased, then looked at your diary and wondered where all those leads actually went, this is what happened.
The report was not lying. It was counting something different from what you were counting, and nobody explained the difference.
What a conversion actually counts
Google Ads records a conversion when someone completes the action you told it to record. Usually that means a form submission, a click on a phone number, or a click on a booking button.
Read that again, because the important word is click.
A phone conversion fires when somebody taps your number. It does not know whether the call connected, whether you answered, whether it was a customer or a supplier, or whether the person hung up after four seconds because they meant to ring the garage next door.
The platform is not being dishonest. It is measuring the only thing it can see. The problem is that the number then appears in a report headed “leads”, and leads is not what it counted.
The five places the gap comes from
One person counted twice. Someone taps your number, does not get through, then fills in the form. That is two conversions and one human being. Google’s conversion counting setting decides whether repeat actions from the same click are counted once or every time, and the default is not always the one you want for lead generation.
Calls that were never really calls. Wrong numbers, existing customers ringing about an invoice, suppliers, and people who tapped by accident while scrolling. All conversions, none of them new work.
Price shoppers. They rang, asked what something costs, said they would think about it, and rang three of your competitors next. A genuine enquiry, a real conversion, and not a job.
Calls you missed. This one hurts, because it counts as a conversion and produces nothing. If a meaningful share of your calls arrive while you are under a car or up a ladder, your reported conversions and your booked work will never agree.
Quoted and lost. You saw them, priced it, and did not win it. Everything worked and there is still no job.
Every one of those is normal. The failure is not that the gap exists. It is reporting the top number as though it were the bottom one.
What it looks like on a real account
Here are figures from an account we run for an independent garage in Medway. Published with their permission.
In one week the account recorded 32 conversions across two campaigns, at a cost per conversion of about £40 on the main campaign.
In the same week the business took roughly 10 to 15 bookings in total, of which about 60 to 70% came from the ads. That puts ad-sourced bookings somewhere between six and ten.
So somewhere between a quarter and a third of conversions turned into work.
Run the money through it and the picture changes completely. The reported cost per conversion is around £40. The real cost of a booked job, across roughly £807 of weekly spend, is closer to £80 to £135.
Both numbers are true. Only one of them is a business number.
We publish that as a range rather than a single figure because it is a range, and we are still tightening the conversion to booking rate with the client. We would rather show a defensible range with the working attached than a precise figure we cannot stand behind. That, incidentally, is a reasonable test to apply to any agency.
And to be clear about direction of travel: at that cost, against what a job of this type is worth, the campaign is performing well. The point is not that the account is bad. It is that £40 and £135 lead to different decisions, and only one of them describes reality.
Why this rarely gets corrected
Partly inertia. Conversion tracking is set up once, at the start, usually by whoever built the site, and then nobody revisits it.
Partly because the platform number is easier. It arrives automatically, it requires no conversation with the client about their diary, and it is always the more flattering of the two.
And partly incentive. An agency reporting 32 leads looks better than an agency reporting eight jobs, especially in month three when nothing much has compounded yet. Very few agencies are actively deceiving anyone. Rather more have simply never had a reason to go looking for the smaller number.
How to close the gap
None of this is difficult. It is just rarely done.
Set a minimum call length. Call reporting can be configured so that only calls lasting beyond a threshold count as conversions. Sixty seconds removes most wrong numbers and accidental taps immediately.
Check your conversion counting setting. For lead generation you usually want one conversion per click rather than every conversion, so a single person doing three things is one lead rather than three.
Set real conversion values. Most accounts we inherit have every conversion valued at £1 by default, which means the bidding treats a job worth £1,200 and a general enquiry as identical. Assign values based on what each type is actually worth and the automated bidding starts working towards revenue rather than volume.
Feed the outcomes back in. Google Ads can accept offline conversion imports, so the jobs you actually booked go back into the account and the bidding optimises for work rather than enquiries. This is the single biggest improvement available on most local accounts and it is the least used.
Reconcile once a month. Sit the diary next to the report and count. Ten minutes. If the two numbers are far apart, you now know something you did not know before.
That last one costs nothing and no software is required. It is also, more or less, how we run Google Ads.
What to ask your agency this month
Three questions, and the answers are more revealing than the numbers.
What exactly counts as a conversion on my account? If the answer is vague, that is your answer.
How many of last month’s conversions became booked work? A good agency will either know or will want to find out with you. A poor one will explain why the question is unfair.
What is my cost per booked job, not per lead? This is the number that decides whether your advertising makes money. It is remarkable how rarely it appears in a report.
Send us your last report. Tell us what it says and what your diary says. If the two match, we will tell you your agency is doing a good job. If they do not, we will show you where the gap is.
