Cost per booking is the amount you spend on advertising divided by the number of confirmed bookings that advertising produced, over the same period. If you spent 600 on ads last month and 24 people booked because of them, each booking cost 25. It tells you what one new appointment costs to win.
Last updated:
The formula
Cost per booking = ad spend ÷ confirmed bookings from those ads.
Three things have to line up for the answer to mean anything:
- The same period. Last month’s spend against last month’s bookings, not last month’s spend against this week’s calendar.
- The same source. Only bookings that came from the ads you are measuring. Regulars who rebooked at the front desk don’t count.
- One definition of “booking”. Decide what a confirmed booking is and keep to it.
The numbers in this article are made up to show the arithmetic. They are not results from any client, and they are not benchmarks.
Cost per click, cost per lead and cost per booking
Ad platforms report several “cost per” figures. They sit at different distances from your appointment book.
| What it measures | What it leaves out | |
|---|---|---|
| Cost per click | What it costs to get someone to tap your ad | Whether that person ever contacted you or booked |
| Cost per lead | What it costs to get a message, a form or a call | Whether the lead replied, could afford the service, or booked |
| Cost per booking | What it costs to get a confirmed appointment | Whether the client showed up, and what they spent |
A campaign can have a low cost per click and a high cost per booking at the same time. That happens when the ad attracts people who are curious but not ready to book. Judged by clicks, the campaign looks like a success. Judged by bookings, it is losing money. This is why we build campaigns around the booking and report on it, as described on our Meta Ads page.
What counts as a booking
Businesses confirm bookings in different ways, and the choice changes your number. Pick the one that matches how you really work:
- Deposit paid. The strictest definition. Fewer bookings will be counted, but almost all of them are real.
- Slot confirmed in your booking system. The most common definition, and usually the easiest to track.
- Confirmed by your team. For businesses that book by phone or direct message. It works, but someone has to record it every time.
Whichever you choose, write it down. If the definition changes halfway through the year, the cost per booking before and after can’t be compared.
One step further is the cost per completed appointment, which leaves out no-shows. It is the truest figure, but no-shows are recorded in your own system, not in the ad platform, so you have to work it out by hand.
How to work it out in five steps
Pick a period
A full calendar month is usually easiest. Avoid very short periods: a week with three bookings tells you very little.
Find your ad spend
Take it from the billing section of your ad account, not from memory. Use what you were actually charged.
Count the bookings that came from ads
Use the best source you have: a booking event recorded by the ad platform, a “source” field in your booking system, or a simple log where your team notes how each new client found you. Our guide to tracking bookings from Facebook and Instagram ads covers the options.
Divide
Spend ÷ bookings. Our cost-per-booking calculator does this for you and also shows cost per lead and the share of leads that booked.
Note where each number came from
Write “Meta-reported” or “from our booking system” next to the booking count. The two rarely match exactly, and next month you will want to compare like with like.
What is a good cost per booking?
There is no universal good number, and we won’t invent one. A cost per booking is good when it is comfortably below what a booking is worth to you. To judge that, you need two figures from your own business:
- Average booking value: what a typical new client pays for the appointment.
- What it costs you to deliver: products, the stylist’s or practitioner’s time, card fees, rent for the chair.
A made-up example. Suppose an appointment brings in 120, and delivering it costs you 70. You have 50 left before advertising. A cost per booking of 25 leaves 25 of profit on the first visit. A cost per booking of 60 means you lose 10 on the first visit, and the campaign only pays off if that client comes back.
Repeat visits matter a great deal in appointment-based businesses, and they are a fair reason to accept a higher cost per booking. Be honest with yourself about them, though. Use the rebooking rate you can see in your own records, not a hopeful guess.
Common mistakes
- Counting leads as bookings. A message that says “how much?” is not an appointment.
- Counting every booking. If you divide ad spend by all of the month’s bookings, including regulars and walk-ins, the figure looks wonderful and means nothing.
- Mixing periods. Ads seen in the last days of a month often produce bookings in the next one. Compare several months before drawing a conclusion.
- Trusting one source blindly. Ad platforms estimate. Your booking system records. When they disagree, your own records are the reference.
- Leaving out your own costs. A cost per booking below the booking value is not automatically profitable.
- Judging too early. A handful of bookings can swing the figure wildly. Wait until you have enough to see a pattern.
What brings the figure down
We can’t promise a percentage, and anyone who does is guessing. These are the levers that usually matter, in the order we check them:
- Count correctly first. A surprising share of “high” costs per booking are tracking problems: bookings that happened but were never recorded as coming from ads.
- Tell the ad platform about bookings. A campaign that is optimized for confirmed bookings looks for people who book. One that is optimized for clicks looks for people who click.
- State the price range and the time needed in the ad. People who would never pay your prices stop inquiring, which wastes less of your budget and your time.
- Reply faster. Leads cool quickly. A booking link in the first reply removes a step.
- Advertise the services with room in the calendar. A full day can’t take another booking at any price.
Once you know your cost per booking, the next figure is return on ad spend. The ROAS calculator works it out, and our performance marketing page explains why we only report it from verified revenue.
Who this is for, and who it isn’t
Useful for
- Salons, studios, clinics and other businesses that sell appointments.
- Owners who run ads, or pay someone to, and want one number to judge them by.
- Anyone comparing two campaigns, two platforms or two agencies.
Less useful for
- Online stores. Cost per order and return on ad spend fit better.
- Businesses with only a few bookings a month from ads. The figure will jump around too much to guide decisions.
- Businesses that can’t yet tell which bookings came from ads. Fix the tracking first.
Questions people ask
Is cost per booking the same as cost per acquisition?
It is one kind of cost per acquisition, where the “acquisition” is a confirmed booking. The general term can also mean a lead or a sale, so always ask what is being counted.
Should I include the agency or freelancer fee?
For judging the ads themselves, use ad spend only, so campaigns can be compared. For judging whether advertising is worth it overall, add the fee. Say which version you are using.
My ad platform shows a “cost per result”. Is that my cost per booking?
Only if the “result” the campaign is set to count is a confirmed booking. Often it is a click, a message or a lead. Check which event the campaign is optimized for.
How many bookings do I need before the figure is reliable?
There is no fixed threshold. As a rule of thumb, the more bookings in the period, the steadier the figure. With very few, look at several months together.
Can cost per booking be too low?
Yes, if it comes from discounting so heavily that each appointment loses money, or from attracting clients who never return at full price. Check the figure against booking value and rebooking, not on its own.
Sources
This article uses no external statistics. The formula is simple arithmetic, and every example figure is made up for illustration and labeled as such.