- Manage the spread, not the average. The average improves on its own once the bottom comes up.
- Five sites averaging 79 could be five sites at 79, or a 92 and a 64. Only one of those is a business.
- Track the spread over time. A widening gap during growth is the earliest warning that your standard is not travelling.
Why the average is the wrong number
An estate average is a comfortable number. It moves slowly, it rarely looks alarming, and it can stay flat while your worst site falls off a cliff. It is the number most likely to be reported to a board and least likely to predict a complaint.
Customers do not experience your average. They experience one site, once, and they judge the brand on it. That is why the spread is the operational number and the average is the reporting number.
Calculate the spread, then watch it over time
Take your latest audit scores, subtract the lowest from the highest, and you have the spread. Under ten points is tight. Over twenty and your best and worst sites are delivering visibly different experiences to people who think they are buying the same thing.
The trend matters more than the snapshot. A spread that widens while you are opening sites means the standard is not travelling with the growth, which is the moment to fix the onboarding rather than the sites.
If your estate average is flat but your spread is widening, you do not have a stable business. You have an improving top half carrying a declining bottom half.
Normalise before you compare
Sites differ in size, format and footfall, and comparing raw numbers across them invites a fair objection that kills the whole exercise. Use per-cover, per-square-metre or per-thousand-customers figures for anything volume-driven.
Standards scores usually need no normalising at all, which is part of why they are such a useful comparison: cleanliness, welcome and availability mean the same thing everywhere.
Publish it, carefully
A published ranking works when the scoring is unarguable and the support is real. It fails when the audits are inconsistent or when the bottom sites get a league table and no help, because then you have taught managers to manage the audit rather than the site.
The safest version is to publish the spread and the trend to everyone, and the individual ranking to the people who can act on it.
Frequently asked questions
What is a good spread between sites?
Under ten points on a hundred-point audit is tight for most operators. Over twenty is visible to customers.
How many sites do I need for this to be useful?
Two. With two sites the spread is the entire story, and it is usually the moment an owner first realises the standard lives in their head rather than on paper.
Should I compare sites that opened at different times?
Compare them, but expect new sites to sit below until they ramp. Track them separately for the first ninety days so they do not distort the estate picture.