Academy › Publishing › Analytics
The average customer isn't the one who buys the most from you
By the end, you'll calculate the real value of an average customer, weighted by how many actually repeat, not by your most loyal one.
By Yodai · Published September 6, 2026 · Updated October 12, 2026
If you multiply how many times a year your repeat customer comes back by the margin on one visit, you get a number that sounds great: in an example pilates studio, $168 a year. The problem is that number doesn't describe your average customer. It describes your most loyal one.
If you use that number to decide how much to spend on bringing in new customers, you'll overspend. If, like in that pilates studio, out of every ten people who walk in for the first time most never come back even once, that number tells you nothing about them.
The customer you calculated isn't the one you actually have
At the pilates studio, a drop-in class costs $35, with a 60% margin: $21 clear per visit. The customer who gets hooked comes eight times a year, so their value is $168, the number from the start. But out of every ten people who try a class, six never set foot in the studio again. They stop at that $21 from the first visit, and that's it.
Only four turn into the $168 customer. If you calculate your acquisition budget assuming everyone is worth $168, you're only accounting for four people out of ten. The other six will cost you, because you're treating them as if they'll come back, and they won't.
How to actually weight it
Here's the math: you multiply the value of the repeat customer by the share who actually repeat, then add the value of the one-time customer times the share who stop after one purchase. At the pilates studio: $168 times the 40% who repeat, plus $21 times the 60% who don't. The result is $79.80, call it $80. That's the value of any random new customer, not your most loyal one.
You get that share, how many out of ten come back, from your own appointment book or booking software: how many new names came in last quarter and how many of them returned at least once. You don't need a full report; counting one quarter by hand is enough.
If your business has more than two types of customers with different frequencies, the same idea works with more than two groups. The real value is the sum of each group multiplied by its weight, not the number from the group you like best.
A hair salon might have customers who come in for a color touch-up every three weeks, customers who get a one-off haircut every couple of months, and customers who book a special service once a year. That's three different weights to multiply and add, not one. An auto shop might have four: the one-time repair that never comes back, the twice-a-year checkup, the monthly maintenance customer, and the fleet customer who brings in several vehicles. Four groups, four products to add up.
that 60% stops after the first visit, and out of what you spent to bring them in you only recover what that visit leaves you
The mistake of looking only at the loyal customer
With this number, the real value of any given customer rather than your best one, you now have the basis for setting a limit on what you can afford to spend to bring in a new one without eating into your margin. That's exactly what you'll decide in the next lesson.
- You have the weighted value of any new customer, not just the one who repeats the most
- You know what share of your customers repeat, pulled from your own appointments or bookings
- You know how to extend the weighting to three or more groups if your business has different frequencies
- That weighted number, not the most loyal customer's, is the basis for deciding how much to spend on acquisition
Test yourself
Answer before opening each one. If one gives you trouble, go back to the section that explains it.
1. A hair salon figures that a loyal customer (cut and color every month) brings in $600 a year, but only two out of every ten new customers become loyal regulars. The other eight get a single haircut that brings in $30 and never come back. What's any random new customer really worth?
$144: $600 times the 20% who repeat, plus $30 times the 80% who don't (600×0.2 + 30×0.8). The mistake would be setting the acquisition budget on the loyal customer's $600, when 80% never come close to that value.
2. An auto shop has three types of customers: a one-time repair ($90, doesn't repeat, 50% of new customers), a twice-a-year checkup ($300, 30%), and monthly maintenance ($900, 20%). What number should the shop use to set its spending limit on acquiring a new customer?
$315: the sum of the three weighted groups (90×0.5 + 300×0.3 + 900×0.2). Using the $900 full-maintenance customer figure would mean overpaying for half of new customers, who only come in for a one-time repair.
3. A yoga studio owner doesn't keep any record of how many new clients repeat and how many don't. They want to set an advertising spending limit right away. What should they do first?
Before setting any limit, count a quarter's worth of appointments by hand: how many new names came in and how many returned at least once. Without that share, there's no weighted number to calculate, and any limit set without it risks being based on the loyal customer, since that's the only data on hand.