Staff friendliness varies from branch to branch because it is mostly a product of local conditions rather than local people: how the branch manager behaves on the floor, whether staffing matches the actual rush, how new starters are really trained, how often the team turns over, and whether anyone at the branch hears what customers experience. Two branches with the same brand, the same handbook and the same induction can feel completely different at the counter. Head office rarely sees the gap because the signals it has about friendliness are either produced by the branch or arrive too late to coach anyone.
This article explains where the variance comes from, why it stays invisible to head office, how to measure it across branches without turning the exercise into surveillance, and what to change once the gap is visible.
Friendliness is a condition, not a trait
It is tempting to read a cold branch as a branch that hired the wrong people. Sometimes that is true. Far more often the same people would be warm at a different branch, or at the same branch on a different shift.
A team member who is two people short at 8am, with a queue to the door and a manager hiding in the office, is not going to be friendly. The same person at 3pm, with the rush over and a manager on the floor, usually is. Friendliness is what a reasonably decent person does when the conditions allow it.
That matters for head office because conditions can be changed from a distance and traits cannot. If the diagnosis is "branch 14 has rude staff", the only levers are hiring and firing. If the diagnosis is "branch 14 runs short every weekday morning and the manager does not work the floor", there are several levers and most of them are cheap.
The five things that make one branch friendlier than another
Across a group of any size, the same handful of conditions explain most of the difference between a warm branch and a cold one.
The manager's own behaviour on the floor
Staff do what the manager does, not what the handbook says. A manager who greets regulars by name, steps onto the till in a rush and thanks the team at close sets a standard that no training module can match. A manager who works from the back office sets a different standard, and the team follows it just as faithfully.
This is the single biggest driver, and it is almost invisible to head office, because a manager who avoids the floor is usually excellent at reports.
Staffing against the actual rush
Rotas are built against forecast demand. The forecast is often wrong in the same direction at the same branch for months, and nobody corrects it because the checklist says the shift was covered.
A branch that is routinely one person short at its peak will be curt at its peak. Customers who visit at that time describe the branch as unfriendly. Customers who visit at quieter times do not. Head office sees an average.
How training actually happens
The induction says a new starter shadows an experienced colleague for three shifts. At a well-run branch that happens. At a branch under pressure the new starter is on the till alone within an hour, learning from whoever is nearest, including their habits with customers.
The gap between the written training and the delivered training is largest at exactly the branches that most need the training.
Turnover, and who trains the next person
A branch that loses a team member every month is always training. The people doing the training are the ones who stayed, who may be six weeks in themselves. Standards drift one hire at a time, and the manager, if they notice, has no time to correct it.
High turnover is often a consequence of the first two conditions, which is why it rarely improves on its own.
Whether the branch hears what customers experience
A branch where the team knows what customers say about them, in the customers' own words, behaves differently from one that only hears about it when a review surfaces or an area manager mentions a score. Most front-line staff want to be good at this. They cannot adjust to what they never hear.
This is the one condition head office can change for every branch at once.
Why head office cannot see the difference
None of the signals head office normally has was built to carry friendliness.
A mystery shop is one scripted visit. It records whether the greeting was delivered, not whether a regular customer on a Tuesday morning felt welcome. A checklist has no line for warmth. A manager's report is written by the person whose behaviour on the floor is the biggest variable.
Public reviews do carry friendliness, usually as the word "rude" or "lovely", but there are too few per branch to compare one location with another in a given month, and they arrive after the customer has left and told other people. By the time head office reads "rude" in a review, it has no detail, no name and no way to ask what happened.
So the first time head office learns that branch 14 is cold is often when a cluster of reviews says so, which is late, and which says nothing about why.
How to measure friendliness across branches without surveillance
The way to see the gap is to ask customers at every branch, the same way, and read what comes back the same way. Four principles keep this from becoming a monitoring programme that staff resent and work around.
Use the customer's own words, not a score
A friendliness score of 3.2 against 4.1 tells head office that two branches differ. It does not say how. "Nobody looked up when I walked in" and "the person on the till was clearly new and nobody was helping her" are two different problems with two different fixes, and both would score about the same.
Let the customer type a sentence, or record one if they prefer. Keep name and rating optional. The detail is in the words.
Read every branch the same way
If branch managers read their own messages and summarise them upward, head office is back to comparing manager reports. The messages need to go to head office and be sorted by one reader, so that "staff seemed stressed" at branch 3 is counted the same way as the same comment at branch 40. The original message should stay behind every finding so anyone can check the reading.
Outhentik works this way: messages go to head office, every message is read by the same reader, and a problem raised at more than one branch shows up as one issue on the daily brief rather than as scattered comments.
Compare branches on the same problem, not on a league table
A ranked list of branches by friendliness tells each manager where they sit and nothing about what to do. Comparing branches on a specific repeated complaint ("staff did not acknowledge me on arrival" at branches 7, 14 and 22) tells head office what to coach and where.
Look for what repeats, not the single message
One message about a short-tempered team member is a bad moment. The same thing described by several customers across several weeks at one branch is a condition. Head office should act on conditions and let the branch handle moments. Treating a single message as evidence of anything is how a feedback programme turns into a grievance programme.
What head office can do about it
Once the gap is visible and described in the customer's words, the fixes are mostly operational.
Fix the conditions before the people
Check the rota against the times customers describe the branch as cold. Check whether the manager is on the floor at those times. Check how many of the team are under three months in. In most cases the friendliness problem is sitting inside one of those three answers, and none of them is solved by a conversation about attitude.
Coach from the specific, not the general
"Customers say we are unfriendly" is not coachable. "Three customers this month said nobody acknowledged them while they waited for a table" is. Give the branch manager the specific pattern and the customers' own wording, and ask what they think is behind it. They usually know.
Make the warm branches the training material
Somewhere in the group there is a branch that customers consistently describe as welcoming. Find out what its manager does on the floor, how it handles the rush, and how it trains a new starter, and write that down. It is more credible to a struggling manager than a handbook, because it comes from a peer with the same brand and the same constraints.
Tell the branch what was heard and what changed
The team at the counter should hear what customers said about them, the good as well as the bad, and should hear when head office changed something because of it. A branch that sees a second person added to the morning shift after customers described the queue as unfriendly learns that the code on the counter is a line to someone who acts. That is what keeps staff pointing customers to it.
What not to do
Publish a friendliness league table. It punishes branches for conditions head office set, and it tells nobody what to change.
Discipline anyone on the strength of a single message. One customer's account of one visit is a reason to look, not a reason to act against a named person.
Send messages about staff to the branch manager to "deal with". That turns a head office signal into a branch-level complaint about the people the manager works beside every day, and it removes head office from the picture exactly where it is needed.
Script the warmth. A mandated greeting delivered by a team member who is two people short and unsupported does not read as friendly. Customers notice the difference between a line and a welcome.
Read a quiet branch as a friendly one. A branch that sends no messages may be welcoming, or may have the code tucked behind the till. It should show as unknown until someone checks.
Realistic expectations
How many customers will describe their experience at the counter is unknown until the code is on the counter. Some branches will produce a steady picture within weeks. Others will produce very little, and the first thing head office learns about those branches is that nobody is pointing to the code.
Customers can and will still leave public reviews about staff, warm or otherwise. A direct line to head office gives head office the account and the chance to fix the conditions. It does not sit between the customer and any review platform.
Conditions change slowly. A rota fix shows up in customer messages within a few weeks. A manager who starts working the floor changes a branch in months. Turnover takes longer than that. Head office should expect to see the repeated complaint thin out before it disappears.
Frequently asked questions
Why is one branch so much friendlier than another when they have the same training?
Because training is delivered differently at each branch, and because friendliness depends on conditions the training does not control: whether the manager works the floor, whether staffing matches the rush, how long the team has been in place, and whether anyone hears what customers say. The same handbook produces different results in different conditions.
How can head office measure staff friendliness across branches?
By asking customers at every branch in the same way and reading what they say in the same way, with the customer's own words kept behind every finding. A single score per branch hides the reason for the difference. Repeated specific complaints, compared across branches, show what to change and where.
Is it fair to compare branches on friendliness?
It is fair to compare them on a specific repeated complaint and ask why it appears at some branches and not others. It is not fair to rank branches on a single number when head office sets the rota, the budget and the manager.
Should customer messages about staff go to the branch manager?
The branch manager should see what customers say about their own location so they can act on it. Head office should see it first and see it for every branch, so the manager is not the only filter and a pattern across branches is visible as one thing.
How do you coach staff from customer feedback without blame?
Share the pattern, not the individual message. Use the customer's wording. Ask the manager and team what they think sits behind it before suggesting a fix. Pair every negative pattern with what customers said went well. Change a condition where one is the cause, and say so to the team.
What if staff think the feedback code is there to catch them out?
Then it will not be shown to customers and the programme will go quiet. Tell the team what the messages are used for, show them the positive ones, and make the first visible result a fix to something they have been asking for, such as cover at the rush. Staff believe what they see happen.
Try Outhentik free for 7 days, no credit card required
Outhentik gives head office a direct line from customers at every branch: every message read the same way, a case with an owner for each customer who asks to hear back, and a daily view of which problems repeat and where. For friendliness, that means head office sees which branches customers describe as cold, in the customers' own words, early enough to change the conditions behind it.