Cost per lead is the wrong number: track cost per booked call

It measures the cheapest step. A lead is a name and a phone number. It says nothing about whether the person meant it, can afford it, or will pick up when your team calls.

And it is easy to push down. Broaden the targeting, use a form that fills in details automatically, drop the qualifying questions, and the cost per lead falls. The leads usually get worse at the same time. Your dashboard improves while your calendar stays empty.

The cost has not gone away. It has moved to your front desk or your sales team, who now spend their mornings calling people who do not remember filling in anything. That time is real money, and no ad report shows it.

Judged on cost per lead, channel A wins by a factor of four, and in most accounts it would get next month’s budget. Judged on cost per booked call, channel B is half the price. Judged on revenue, B brings three times as much for the same spend.

Nothing about channel A looked wrong in the ad platform. The leads were cheap, the forms worked, the reports were green. The problem only shows when someone follows the leads through booking, attendance and payment, which is the part most reporting skips.

Your own numbers will differ, sometimes in A’s favour. The point is that you cannot know which channel is better until you track past the lead.

Three things have to be true. Every lead must carry its source, captured when the person first arrives and saved with the lead. Every lead must move through clear stages: new, contacted, booked, attended, won or lost. And someone must record what actually happened, such as the receptionist marking a no-show or the salesperson entering the amount of a closed deal.

Most businesses already have the information. It sits in a spreadsheet, an agenda and a till that do not talk to each other. The work is joining them. Our booking forms save the name and phone at the first step, ask the qualifying questions, book into the team calendar and send the lead to the CRM with its source, which removes most of the joining by hand.

Then report once a month, by source: leads, booked calls, attended, customers, revenue, and spend. Six columns. That table settles most budget arguments.

Booking rate depends mostly on speed and friction. Reply within minutes while the person still remembers you, ideally on the channel they prefer; our guide on WhatsApp follow-up covers this in detail. Let people pick a slot themselves, with no callback to wait for. Ask two or three qualifying questions on the form so your team calls the right people first.

Show rate depends on what happens between the booking and the appointment. Send a confirmation straight away, a reminder the day before and another a few hours before. Make rescheduling easy, because a moved appointment beats a missed one. Call no-shows the same day and offer a new slot.

None of this is advertising, and it often moves revenue more than any change to the ads. That is the reason lead handling is a service of its own.

Ad platforms optimise toward the events you send them. If the only event is a form submission, the algorithm gets very good at finding people who submit forms, including the ones who never answer.

Meta’s Conversions API lets you send events from your own systems, such as a lead marked qualified or an appointment attended, so campaigns can optimise for those. Google Ads accepts offline conversion imports for the same purpose. Our revenue attribution ties each lead and booking to its campaign and sends qualified leads back to Meta this way.

Start with booked calls as the signal. They happen often enough for the platforms to learn from, and they sit much closer to revenue than a form fill.