Everything Tests Fine but the Customer Isn't Satisfied: Decision Tree

Why this matters

You have done the hard part. You reproduced the complaint, measured at the point of use, checked the sensory and adjacent causes, and the system is genuinely delivering. The customer is still not happy. This is where good techs lose jobs, because they treat "in spec" as the end of the conversation when it is really the start of a different one. Nothing here is a repair. Everything here is a routing decision.

Start here: confirm you actually finished the diagnosis

Do not enter this tree on a hunch. You belong here only if all three hold:

• The complaint is not a hazard cue. If a smell near combustion, a shock, warmth at electrical, or symptoms that ease when they leave the house are in play, leave this tree and handle the hazard first: ventilate and shut down, or de-energize and verify dead.
• You measured at the complaint point and time, not just at the equipment.
• You either reproduced it in spec or confirmed it is not reproducible.

If any of those is missing, go back to the measurement tree. Routing a resolution before the diagnosis is done is how you promise the wrong fix.

Is there room to move within spec?

Being in spec is a range, not a single point. Often the system can be nudged toward the complaint without exceeding any limit.

• If a setting, a balance, a schedule, or a level can shift within the allowed range toward what they want: make the adjustment. This is the cheapest, cleanest resolution, and it is real work.
• If the system is already at the edge of its range in the direction they want: you cannot tune your way out. Go to the next branch.

Is the complaint about understanding, not performance?

Sometimes the system does exactly what its settings tell it to, and the customer does not know that, or is fighting the settings.

• If correct operation is being misread as a fault (a setback they forgot, a normal noise, an expected recovery time): the fix is education, not a change. Show them, set it with them, write it down.

Is it a preference for different behavior?

The system works correctly and they simply want it to work differently than designed.

• If they want behavior outside how it was built or set, and nothing is defective: this is a change, not a repair. Scope it, price it, and let them decide. A preference is legitimate, and meeting it is billable work, not a warranty callback.

Is it a real limit of the design?

The system is doing everything it can, and everything it can is not enough for this space or this use.

• If the design itself is the ceiling (a run too long, a stage too coarse, a fixture too far, capacity too small for the load): the honest answer is a design conversation. Live with the limit, change habits, or upgrade the design. Do not sell an upgrade to dodge the honesty; name the limit plainly first.

Still stuck? Get a second set of eyes

If none of the routes fit and the customer is genuinely experiencing something you cannot name, that is a signal, not a defeat.

• Bring a second tech or a specialist, or offer to observe under the exact conditions that trigger it. Sometimes the missed variable only shows up to fresh eyes.

Recap

1. Confirm no hazard and that the diagnosis is truly complete.
2. Tune within spec if there is room.
3. If it is a misunderstanding, educate.
4. If it is a want, not a defect, scope it as a change.
5. If the design is the ceiling, have the honest options conversation.
6. If nothing fits, escalate to a second opinion rather than force a fix.

References

• Trade-standard practice on performance verification and acceptance criteria
• Manufacturer documentation on rated capacity and design limits
• See related: Chase a Comfort Complaint Further or Explain the Limits (decision tree); When the Fix Is Education, Not a Repair; Diagnosing a Complaint That's Really About Preference