A twelve-dollar lead who never walks through the door is more expensive than a forty-five dollar lead who books and buys. Everyone agrees with this sentence and almost nobody runs their account as though they believe it.
Why the number is so seductive
Cost per lead is available immediately, it sits at the top of every dashboard, and it moves. You can change it this afternoon. Broaden the audience, loosen the form, shift budget to a cheaper placement, and the number goes down before dinner. It feels like progress because something visibly improved.
Every other number in the chain takes weeks and requires talking to someone in operations.
What the platform is actually doing
If your campaign objective is set to leads and your conversion event is a form submission, the platform has been given one instruction: find people likely to submit a form. It will do that superbly. It will find the people most inclined to tap a button, which is not the same population as people inclined to spend two thousand dollars on a treatment package.
Bad lead quality is almost never a targeting failure. It is the account doing precisely what it was told, toward a finish line nobody moved.
The chain the number hides
Between an impression and money there are eight steps, and cost per lead reports on three of them.
- Contact — did anybody actually reach this person, and how long did it take?
- Booked — did the conversation produce an appointment?
- Confirmed — did anything happen between booking and the day?
- Arrived — did they turn up?
- Sold — did they buy?
A clinic can halve its cost per lead and lose money doing it, if the leads it gained convert at a third of the rate of the ones it replaced. This happens constantly and is nearly invisible, because the metric that got worse lives in a different system from the metric that got better.
What to look at instead
Cost per arrived customer. Cost per sale. Those are the two numbers that decide whether a campaign is worth running, and both require lead source to survive the journey into whatever system holds your bookings.
If that connection doesn't exist, building it is more valuable than any optimisation you could make this quarter — because until it does, every budget decision you make is a guess wearing a spreadsheet.
One diagnostic worth running this week
Take your two best-performing ads by cost per lead. Then find out, for each, how many of those leads actually arrived. In my experience two ads at an identical cost per lead routinely produce completely different downstream conversion, and the gap between them is the most useful thing in the account.
You can optimise toward cheap, or you can optimise toward revenue. The platform will do whichever you ask, and it will never tell you that you asked for the wrong one.
None of this means cost per lead is useless. It's a fine input. It is a terrible objective.