Symptom Resolved on Arrival Leave vs Investigate vs Monitor Decision Tree

Why this matters

You arrive and the problem is gone. The system runs fine, the noise stopped, the leak is dry. This is one of the most error-prone moments in field service. Pack up and leave and you may have missed an intermittent fault that returns the next day, earning a frustrated callback and a customer who feels you did nothing. Tear into a full invasive diagnostic on a system that is currently healthy and you may create the very fault you came to find, or charge for a hunt with no target. The right move depends on what kind of fault hides behind a disappearing symptom, and reading that correctly separates a diagnostician from a parts-swapper.

The situation

The reported symptom is not present when you arrive. The customer described something real, but the system is behaving now. You must decide whether to leave with a sound explanation, investigate to find a root cause that is hiding, or set up monitoring to catch the fault when it recurs, all without manufacturing a problem or dismissing a real one.

Decision factors

• Symptom severity if it returns. A disappearing symptom that signals a safety or catastrophic-failure risk (intermittent electrical fault, sporadic gas smell) demands investigation even when absent. A minor nuisance does not.
• Reproducibility. Can you recreate the conditions that produced it, load, temperature, runtime, cycle stage? If you can force it back, investigate. If it depends on conditions you cannot stage, monitoring is the realistic tool.
• Explainability. Is there a clear, benign reason it resolved (a tripped protection reset itself, ice melted, air purged)? A satisfying explanation supports leaving. An unexplained disappearance does not.
• Pattern. Is this the first occurrence or a repeating intermittent? A recurring intermittent fault is real and worsening, not a fluke, and earns deeper work.
• Evidence on the system. Are there fault codes, residue, wear marks, scorching, water staining that corroborate the customer's report even though the symptom is absent? Physical evidence redirects you from "nothing wrong" to "find the cause."
• Cost of being wrong. Weigh the callback and trust cost of leaving prematurely against the cost and risk of invasive work on a healthy system.

The decision

Investigate now when the symptom, if real, carries safety or major-failure risk; when you can reproduce the triggering conditions; when stored fault codes or physical evidence point to a cause; or when this is a repeat intermittent rather than a first event. Bounded, evidence-led investigation of a hiding fault is real diagnostic value. Follow the evidence, not a hunch, and avoid invasive teardown of a healthy system without a lead.

Monitor when the symptom is condition-dependent in a way you cannot stage on the spot, there is no immediate safety risk, and there is no current evidence to chase. Set up the means to capture the fault on recurrence: enable or check fault logging, leave or place data-logging where feasible, instruct the customer precisely on what to note and when to call, and define the trigger for a return visit. Monitoring converts an absent fault into a future caught fault instead of a guess.

Leave with an explanation when there is a clear benign cause for both the symptom and its resolution, no safety risk, no corroborating evidence, and no recurrence pattern. Explain to the customer what likely happened and why it resolved, document it, and define what would warrant calling back. Leaving is legitimate only when you can name why the symptom appeared and disappeared, not merely because it is quiet now.

Never swap parts to "cover" an absent symptom you cannot diagnose. Replacing components on a healthy system without evidence is guessing on the customer's dime and can introduce new faults.

The trap to name out loud is the resolved-but-unexplained symptom. A fault that vanished for no reason you can identify has not been fixed; it has gone quiet. Intermittent electrical faults, marginal connections, thermal expansion issues, and early-stage mechanical wear all produce symptoms that come and go on their own schedule. Treating "it works now" as "it is fine" is how a tech earns a callback and a customer who feels unheard. The discipline is to require an explanation before you classify a disappearance as benign: a protection device that reset, a frozen line that thawed, trapped air that purged, a temporary condition that passed. If you cannot name the mechanism, you do not have a benign resolution, you have an absent fault that needs monitoring.

Calibrate investigation depth to safety, not to billability. A disappearing symptom that could indicate an electrical, combustion, or structural hazard earns investigation even when the system runs clean and even when the customer would rather you just leave. The cost of missing a recurring safety fault dwarfs the cost of a thorough check, and "it was working when I left" is not a defense if the fault was the kind that kills intermittently.

What to document

• The reported symptom and that it was absent on arrival.
• Conditions you tested or could not stage, and the results.
• Any fault codes, physical evidence, or recurrence pattern found or ruled out.
• The decision (investigate, monitor, leave) and the reasoning.
• For monitoring: the logging set up, the customer instructions given, and the return trigger.
• For leaving: the benign explanation and the call-back conditions.

References

• Air Conditioning Contractors of America (ACCA), diagnostic-procedure and intermittent-fault guidance in member technical and quality-assurance materials (acca.org).
• NFPA 70B, Standard for Electrical Equipment Maintenance, on condition assessment, fault logging, and investigating intermittent electrical faults.
• Plumbing-Heating-Cooling Contractors Association (PHCC), troubleshooting and service-documentation resources on no-fault-present visits (phccweb.org).
• Occupational Safety and Health Administration, guidance on recognizing intermittent electrical and gas hazards that may be absent at the moment of inspection (osha.gov).