The booking page is the last thing they touched before becoming a customer.

Somebody booking an appointment types a real email address because they want the confirmation, and a real phone number because they expect a reminder.

That makes the contact details on a booking unusually accurate — better than a contact form, where a wrong address costs the person nothing.

It also carries a timestamp and an appointment type, which is more structure than most lead files have.

The result is a high match rate against a customer record, which means less revenue disappearing into Direct / Unknown for want of a usable key.

Most businesses running a scheduler already have months of this and have never exported it once.

A booked appointment that nobody attends has consumed a slot somebody else could have had, and for a practitioner that slot is the entire inventory.

Channels differ substantially in how often their bookings are kept: an appointment booked after a long consideration is attended far more reliably than one booked on an impulse from an advert.

Counting only the appointments that happened hides that completely and makes the impulsive channel look efficient.

Keeping cancellations and no-shows in the file gives a real cost per attended appointment per source, which is the number a practitioner's diary actually runs on.

It is also frequently a reminder-settings finding rather than a channel one, which is a cheaper thing to fix.

A scheduler sits at the end of the journey, so the referrer it sees is usually your own website rather than whatever produced the visit.

A single free-text or dropdown question — how did you hear about us — captures what no pixel can: the neighbour, the podcast, the sign, the search somebody made a month ago.

It is self-reported and therefore imperfect: people misremember, and the last thing they saw is not always what decided them.

So the report keeps it as its own labelled source beside any tracked ones rather than merging them, because two imperfect signals reported separately beat one blended figure that looks precise.

For a small service business it is often the only source data that exists, and it is better than nothing by a very wide margin.

A free fifteen-minute consultation and a paid ninety-minute session are different products with different values and different buyers.

A channel producing mostly one will be ranked as though it produced the mix, which is how a good source of paid work gets cut for producing few bookings.

Splitting on appointment type is immediate, requires nothing extra, and usually reveals that one type produces nearly all the revenue.

That changes where the booking links go on the website, which is a cheaper intervention than changing the media plan, and it can be tested within a week rather than a quarter.