The service recovery mistakes that cost multi-location operators most are made at group level, before any phone call happens: building a callback process when the only way a customer can reach you is a public review; treating recovery as a branch skill rather than a group system; fixing the customer and leaving the cause in place at every other branch; letting the remedy decide the outcome; scaling by adding area managers to an inbox; keeping recovery records where head office cannot compare branches; closing the loop with the customer but not with the branch; reading silence after a callback as satisfaction; and recovering in a language the customer did not write in. Each has a cheap fix, and most of the fixes are decisions rather than tools.
This article takes the nine in turn, with what to do instead. It assumes the basics are in place: a named owner for each case, a small set of states, and a visible clock. If they are not, start with those and come back.
1. Building the callback process before the inbound line exists
Many groups write a recovery standard, train managers on the call, and then find there is nothing to recover from, because the only way a customer can reach the business is a review on a public platform or a complaint at the counter. Reviews arrive late, with no contact details and no way to reply privately. Counter complaints come from the small number of customers angry enough to raise them in the moment.
What to do instead: open the inbound line first. A code at the counter, on the table or on the receipt, or a link sent after the visit, with nothing required of the customer and a choice to be contacted. Then build the callback process around what arrives. Volume is unknown until the line is open, and that is the first thing a rollout finds out.
2. Treating recovery as a branch skill rather than a group system
In a group where each branch handles its own complaints its own way, the good branches recover well, the weak branches recover badly, and head office cannot see which is which. A customer's experience of the group depends on which branch they happened to complain at.
What to do instead: one process, with the same states and the same clock at every branch, and every case visible to head office. The branch still makes the call. What is shared is the structure and the view, so that a branch that is slow to contact customers is visible as slow rather than as quiet.
3. Fixing the customer and leaving the cause in place
A customer at branch 9 is called back about a cold meal, offered a replacement and marked resolved. The same thing is happening at branches 4 and 17, where nobody connected the three cases, because each was handled as a single callback. The group fixed three customers and left the kitchen process alone.
What to do instead: read every message the same way at every branch, so the same complaint at several locations shows as one issue rather than as unrelated cases. Recovery handles the customer. The issue list handles the cause. They are different lists with different owners, and a group needs both.
4. Letting the remedy decide the outcome
Two failure modes, usually both in the same group. At some branches the refund is the default because it is the fastest way to close the case, and the remedy budget quietly becomes the recovery process. At others nobody is sure what they are allowed to offer, so they offer nothing and the case sits open while someone asks.
What to do instead: set remedy authority in writing, by role and by amount, and make it generous enough that the person on the call does not need to escalate for the ordinary case. Then measure reopened cases, not refund totals. A branch with low refunds and many reopened cases is not saving money.
5. Scaling by adding people to the inbox
When volume rises, the instinct is to add area managers to the shared inbox. Ownership diffuses, the area manager covering twelve sites becomes the slowest link, and cases wait for a person who is on the road three days a week.
What to do instead: ownership is one person per case, assigned at head office, and the owner is usually at the branch. The area manager is an escalation path and a reader, not a default owner. If head office cannot assign cases to branches within a day, that is the capacity problem to solve, not the inbox.
6. Keeping the records where head office cannot compare them
Recovery records in a spreadsheet per branch, a WhatsApp group per region, or the branch manager's notebook. Each is adequate for that branch. None of them lets head office see the oldest open case across the group, compare time to first contact between branches, or notice that one location reopens cases at twice the rate of the others.
What to do instead: one record of cases, with the same fields at every branch, that head office and the branch both see. The branch sees its own. Head office sees all of them. Comparison between branches is the point, and it is impossible when the records live in different places.
7. Closing the loop with the customer but not with the branch
The callback happens, the customer is satisfied, the case is marked resolved at head office, and the team at the branch never hears that the complaint they were worried about was closed well, or what was learned. Over time the team sees complaints go up and nothing come back down.
What to do instead: the branch manager sees every case about their own location, including its resolution, and tells the team what happened. Where a case led to a change (a rota, a process, a piece of equipment), head office says so to the branch. Recovery is partly for the customer and partly for the team's belief that the channel leads somewhere.
8. Reading silence after the callback as satisfaction
A case is marked resolved after the call. The customer does not get back in touch. The group reads that as success. It may be, or the customer may have decided the follow-up was not worth their time and simply not returned.
What to do instead: resolved should mean the agreed action was carried out, not that the call happened. Where the fix takes time, the case stays open until it is done. Keep a reopen path that the customer can use without starting again, and treat reopened cases as the most informative number recovery produces.
9. Recovering in a language the customer did not write in
In a market where customers write in more than one language, a message in Arabic read by someone who works in English, or the reverse, loses detail and tone. The callback happens in the wrong language or with the wrong understanding of what was said, and the customer's second message is sharper than the first.
What to do instead: read every message in the language it was written in, with the same scheme in both languages so branches compare fairly, and call back in the customer's language. Where that is not possible at the branch, head office assigns the case to someone who can.
How these mistakes connect
Most of the nine are one mistake in different forms: recovery set up as a branch activity that head office hears about afterwards, rather than a group system head office sees as it happens. Opening the inbound line to head office, one record of cases, one way of reading messages, and the branch seeing its own cases with their outcomes removes most of the list at once.
Outhentik is built on that shape: the customer's message goes to head office, a customer who asks to hear back becomes a case with an owner, a state and a clock, a problem raised at more than one branch becomes an issue on the daily brief, and the branch manager sees everything about their own location and nothing about anyone else's.
What to expect
Opening the line changes what arrives before it changes how fast anything is fixed. The first weeks mostly reveal which branches show the code and which do not.
Reopened cases rise before they fall. Once resolved means the fix was done rather than the call was made, cases that would previously have been closed are kept open or come back. That is the measure working.
Remedy spend goes up at some branches and down at others as authority is set in writing. The group total usually matters less than whether reopened cases fall at the branches where spend changed.
None of this removes public reviews. Customers who can reach head office directly can still post wherever they like. Recovery gives the business the account and the chance to fix the problem. It is not a way to manage what customers say elsewhere, and should not be presented internally as one.
Frequently asked questions
What is the most common service recovery mistake in multi-location businesses?
Setting recovery up as something each branch does its own way, with records head office cannot compare. The weak branches recover badly and nobody can see which ones they are. The fix is one process, one record of cases, and every case visible to head office with the branch seeing its own.
Should the branch or head office own customer callbacks?
The call usually belongs at the branch, because the branch can fix what went wrong. Assignment and visibility belong at head office, so no case depends on the branch that caused it choosing to act, and so the oldest open case in the group is visible to someone.
How do you stop refunds becoming the default remedy?
Set remedy authority by role and amount, make it generous enough for ordinary cases, and measure reopened cases rather than refund totals. A refund that closes a case the customer later reopens was not a saving.
When should a case be marked resolved?
When the agreed action has been carried out, not when the call has been made. If the fix takes a week, the case stays open for a week. A customer who comes back after resolution reopens the case rather than starting a new one.
How do you find problems that affect more than one branch?
By reading every customer message the same way at every branch, so the same complaint at two or more locations appears as one issue. Handling each case on its own, branch by branch, hides the pattern until it is large.
Does good service recovery reduce negative reviews?
That is not what it is for, and it should not be promised. Recovery gives the business the customer's account and the chance to put the problem right. What the customer says publicly afterwards is their decision, and nothing in a recovery process should sit between them and it.
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. Most of the mistakes in this article come from recovery that head office hears about afterwards instead of seeing as it happens.