top of page

Handling callbacks and warranty work without losing the customer

  • Aug 27
  • 3 min read

Updated: 17 hours ago

Introduction


A callback is a truck roll nobody pays for. The technician's time, the fuel, the slot a paying job could have used — all spent returning to work you already invoiced.

It is also the single cleanest quality signal a trade business has, and almost nobody measures it. Most owners have a vague sense that callbacks happen sometimes, which is not enough to manage anything.


1. Handling callbacks and warranty work starts with counting them


Define a callback precisely: a return visit to the same property for the same fault within a set window, say ninety days.

Then count them as a percentage of jobs completed, by technician and by job type. The rate is almost always higher than the owner assumed, and the distribution is almost never even — a small number of job types or one technician usually accounts for most of it.


2. Go back quickly, and go back free


Whatever caused it, the customer's experience is that they paid you and the problem returned.

Attend fast and do not charge. Arguing about whether it is technically covered, in the customer's hallway, costs more in reputation than the visit costs in labour. The customer who was looked after on a callback often becomes more loyal than one whose job went smoothly, because they have now seen what you do when something goes wrong.


3. Never let the technician debate fault at the door


The person who did the original work is the worst-placed person to adjudicate whether it was their mistake.

Give them a standing instruction: fix it, document it, and let the office review afterwards. Any conversation about liability happens later, internally. This protects the customer relationship and it protects the technician from a defensive argument they cannot win.


4. Separate the three real causes


Callbacks look identical from outside and have completely different fixes.

There is workmanship, where the work was not done correctly. There is diagnosis, where the fault found was not the fault causing the problem. And there is expectation, where the work was correct but the customer understood something different. Recording which one it was is what makes the data useful.


5. Attack diagnosis errors first


Workmanship failures are usually rare and obvious. Diagnosis failures are the expensive, repeating ones.

They happen when a technician treats a symptom under time pressure, or replaces the component that commonly fails without confirming it. The fix is process — a required confirmation step before replacing anything, and permission to take the extra twenty minutes.


6. Prevent the expectation callbacks with a closing conversation


A large share of callbacks are jobs that were done correctly.

The technician fixed what was agreed, and the customer expected a different or broader outcome. Two minutes at the end — what was done, what was not done, what to expect, what would warrant calling back — removes most of these. It is the cheapest intervention available here.


7. Write the warranty down before you need it


An unwritten warranty is whatever the customer believes it to be.

State the period, what is covered, what is excluded, and what happens if the same fault recurs. Put it on the invoice. Customers find a clear, modest warranty more reassuring than a vague generous one, and it gives your office a document to work from rather than a memory of a conversation.


8. Feed it back to the technicians without blame


A callback rate published as a league table produces hidden callbacks, not fewer of them.

Review them as cases: what was found, what was actually wrong, what would have caught it. Technicians will surface their own callbacks if doing so is treated as information rather than as a disciplinary event, and you cannot fix what nobody reports.


9. Put callback rate on the monthly numbers


Next to revenue and average ticket.

It is the best single proxy for whether the work is actually good, it predicts review scores, and it directly measures margin that is leaking out of completed jobs. A business that tracks it for six months usually finds two or three specific process fixes worth more than any marketing change available.


Conclusion


Define a callback precisely and count it by technician and job type, because the rate is higher and less evenly spread than owners assume. Attend fast, attend free, and forbid fault debates at the door.

Separate workmanship, diagnosis and expectation causes, attack diagnosis errors with a required confirmation step, prevent expectation callbacks with a two-minute closing conversation, write the warranty on the invoice, review callbacks as cases rather than as a league table, and keep callback rate on the monthly numbers.


Related reading


 
 
 

Comments


bottom of page