Your Own Repair Didn't Hold: Decision Tree

Why this matters

This is the callback that stings: same customer, same complaint, your invoice on the last visit. The equipment is not the only thing being tested now - so is whether you can find your own miss without your ego steering you back to the same wrong answer. The general callback tree asks "whose fault and what cause." This one starts from an established fact - it was your repair - and walks the two things that decision adds: honestly classifying which of your own errors it was, and holding the customer's trust while you do it. Get both right and a failed repair becomes the reason they keep calling you.

Start here: is the recurrence a hazard

If the fault came back as something dangerous, safety leads.

• If it returned as a gas smell, smoke, a burning odor, arcing, water at live electrical, or a pressure or structural risk: handle the hazard first - get people clear, de-energize or shut off from a safe position, call the utility if it is gas. Diagnose the callback only after the danger is controlled. A comeback does not lower the safety bar; it often raises it, because something is failing repeatedly.
• If it is not a hazard, work the tree below.

Step one: own it before you open anything

How you arrive sets whether the diagnosis happens in a trusting room or a hostile one.

• Say it plainly: "It's back, I'll find out why, and we'll make it right." Do not defend the first repair before you have re-checked it.
• Defending your first call anchors you to it, and that anchor is what hides the miss. Arrive open or you will re-confirm your own error.

Step two: same complaint, or a different symptom

• If the symptom is different from the original, treat it as a possible new or second fault, not automatically your failed repair. Verify it before you either claim it as new or eat it as yours.
• If it is the same complaint returning, it is a true comeback on your work. Continue down the self-classification below.

Step three: classify which of your own errors it was

A repeat of the same fault is one of three misses. The tell tells you which.

• Wrong diagnosis. The part you replaced was not the cause. Tell: the thing you fixed is still fine, and the actual driver was never touched. You solved a problem the unit did not have.
• Right target, symptom only. You replaced the real casualty but left the disease. Tell: the same wear part failed again the same way, fast. Something upstream is still killing it - an over-current, a restriction, a contamination, a misalignment.
• Right diagnosis, bad execution. Your call was correct; the repair did not hold. Tell: a connection you made is loose, a fastener under-torqued, a part not seated, or a close-but-wrong replacement that ran briefly then quit under real load.

Re-run the diagnosis from the symptom forward, not from the part you touched. The miss is almost always the branch your first visit talked itself out of.

Step four: rule in a genuinely separate fault, honestly

Before you decide it is all on you, test the one path that is not.

• If your original repair checks out sound - the part is good, the joint is tight, the cause you named is addressed - and a different component or a masked second problem is producing the symptom, this is a new fault, not a failed repair.
• Prove it to yourself with a reading before you say it to the customer. The temptation to relabel a miss as a "new issue" is strongest exactly here, and getting caught shading it costs the relationship.

Step five: attribution decides who eats it

• Your miss - wrong diagnosis, symptom-only, or bad execution - you cover, and you say so early and without being asked. A callback you eat graciously buys more loyalty than the first job did.
• A proven separate fault is a new job, but lead with the evidence of what you fixed correctly and why this is different, before any talk of a new charge.

Step six: close it so it does not come back a third time

• Fix the root cause, then prove it under the exact failure condition before you pack up - same load, same heat, same point in the cycle.
• Record the reading that confirms it, so a future visit has a baseline instead of a memory.
• Document what differed from the first visit. A third trip on the same fault is the one that ends the relationship; a documented root-cause close is what prevents it.

Recap

1. If it came back as a hazard, secure it before diagnosing.
2. Own it at the door; do not defend the first repair.
3. Same complaint means classify your miss: wrong diagnosis, symptom-only, or bad execution.
4. Rule in a truly separate fault only with a reading, not a hope.
5. Eat your miss early; prove a new fault before you bill it.
6. Fix root cause, prove it under the failure condition, and document what changed.

References

• Trade-standard root-cause analysis and callback-handling practice
• See related: When the Fault You Already Fixed Comes Back
• See related: Callback: The Fix Didn't Hold; Two Independent Faults Masking Each Other