Callback: The Fix Didn't Hold

Why this matters

A callback costs you twice: the return trip you do not bill, and the customer's confidence. Worse, most callbacks happen because the first visit treated a symptom and missed the cause. The discipline that prevents the third trip is refusing to repeat the same repair. If you fixed it once and it broke again, the fix was wrong, incomplete, or aimed at the wrong target. Diagnose the callback, do not just redo the work.

Start here: was it ever actually fixed

Before anything else, confirm the original complaint and the new complaint are the same fault.

• If the customer reports a different symptom than the original, treat it as a new job. It may be unrelated, or your repair may have exposed a second pre-existing problem (see the related article on one fix revealing another fault).
• If it is the same symptom returning, continue down this tree. This is a true callback and the first fix did not hold.

Step 1: did the repair address the cause or the symptom

This is the most common reason a fix fails. The first visit replaced the part that failed without asking why it failed.

• If you replaced a consumable or wear part (a fuse, a belt, a seal, a filter) and it failed again fast, something is destroying it. An over-current, a misalignment, a contamination source, an over-pressure. Find what is killing the part before you fit a third one.
• If you adjusted a setting and it drifted back, something is moving it - vibration, thermal cycling, a loose fastener, or a control that is fighting the setting. Lock it or fix what moves it.

A part that fails twice the same way is a flashing sign that the root cause is still live.

Step 2: was the repair done correctly

Rule out workmanship before you go deeper.

• Check your own connections, torque, and seating. A connection that was loose, a fastener under-torqued, a part not fully seated will work for a day and then quit. Re-check every joint you touched. (See the related article on why torque matters.)
• Confirm you used the right part. A close-but-wrong replacement - wrong rating, wrong size, wrong spec - can run briefly then fail under real load. Compare what you installed against the nameplate.
• Verify you reassembled in the right order. A skipped step, a reversed orientation, or a missing piece will hold only until the system is fully exercised.

Step 3: is the real fault somewhere you did not look

If the part was right and the work was clean, the cause is upstream or adjacent.

• If the failed part is downstream of another component, test the feeder. A bad supply, an intermittent control, or an upstream restriction will take out the part you keep replacing.
• If the fault is intermittent, it likely did not reproduce on the first visit, so you fixed what you could see and the hidden fault returned. Reproduce the failure condition this time - same heat, same load, same point in the cycle - before you call it fixed.

Step 4: did the system or the conditions change

Sometimes the first fix was correct and something else moved.

• If conditions shifted since the first visit (weather turned, load increased, a setting was changed by the customer), the system may now be exposing a different weak point. Confirm what changed since you left.
• If the customer operated it differently, a usage pattern you did not account for can defeat an otherwise sound repair.

Closing the callback the right way

When you find the true cause, fix that, then prove it under the failure condition before you leave. Run the system through the exact scenario that produced the complaint. Write down the readings so the next person - or you on a future visit - has a baseline. A callback you diagnose to root cause is the last trip. A callback you patch again is the second of three.

References

• Trade-standard root-cause analysis practice
• Component manufacturer specifications (for correct-part verification)
• See related: Fixed One Thing, Broke Another
• See related: Torque: Why It Matters and Getting It Right