Documenting Your Findings When a Third Party Will Review Them

Why this matters

Most service notes are written for your own shop and never read by anyone else. A finding that a third party will review is a different animal, because the reader is not a filing clerk, it is a technical peer who may be paid to disagree with you and who can re-test what you wrote. Documentation that would pass an internal read falls apart the moment an adversarial expert picks it up looking for a gap. This card is about writing a record that survives that read, not just one that records what happened.

Assume an adversarial re-reader, not a friendly one

Write every line as if the person reading it wants to knock it down. That single assumption changes what you put on the page:

• A friendly reader forgives a shorthand conclusion. An adversarial one treats any conclusion without evidence as an opinion and discards it.
• A friendly reader trusts your number. An adversarial one asks what you measured it with and against what reference.
• A friendly reader assumes you checked the obvious exclusions. An adversarial one assumes you did not, unless you wrote that you did.

If your note only makes sense to someone who already agrees with you, it is not documentation, it is a memo to yourself.

Separate observation from inference on the page

The most common way a finding gets torn apart is that observation and conclusion are welded into one sentence, so disputing the conclusion looks like disputing the fact. Split them physically:

• Observation is what a camera or a meter would capture: "Terminal reads open. No discoloration or contamination present."
• Inference is what you concluded: "Consistent with an internal break rather than an external cause."

A reviewer can argue your inference, which is fair, but a clean observation stands until they re-test it themselves. Welded together, your whole finding falls when they push on the soft half.

Make every finding independently reproducible

The gold standard for a contested finding is that the reviewer can repeat your test and get your result. Give them what they need to do that:

• State the conditions. What was energized or running, what the ambient was, what setting the unit was on. A reading with no conditions is a reading nobody can reproduce.
• State the instrument and the reference. The reading, what took it, and the spec or baseline you compared it to, side by side.
• State the access. What you could see directly versus what you inferred was behind a wall, under insulation, or inside a sealed assembly. Never let a guess wear the clothes of an observation.

A finding built to be reproduced is far harder to dismiss than one built to be believed.

Preserve the evidence in a re-testable state

Paper is not the only record. The failed part is evidence, and destroying or discarding it hands the other expert the last word.

• Do not throw out the failed component in a disputed case. Bag it, label it, and note where it is. A reviewer who can re-examine the actual part cannot invent a different one.
• Photograph before you disturb anything, because a repair erases the scene the same way a wipe erases corrosion.
• If you must alter the equipment to make it safe, document its state before and after, so the change reads as a safety action, not as tampering.

Use neutral language a hostile reviewer cannot weaponize

Loaded words become ammunition. "Obviously neglected" and "clearly a factory defect" are conclusions dressed as observations, and a sharp reviewer uses that overreach to paint your whole finding as biased.

• Describe, do not characterize. "Dust and debris packed in the coil, consistent with extended service without cleaning" is stronger than "badly neglected."
• Avoid naming fault or motive in the record. State the physical evidence and let attribution follow from it. See related: Separating a Technical Fact From a Financial Motive in a Dispute.
• Do not editorialize about the other expert. A jab in the file makes you look like the party with an axe to grind.

Distribute the same record to all parties at once

In a three-way dispute, the fastest way to lose credibility is to have told slightly different stories to the customer, the inspector, and your office. Write one record and give the same one to everyone involved on the same day. When all three parties hold identical documentation, the "he said something different to me" version of the dispute cannot start. It also forces the other expert to argue against a fixed record instead of a moving one.

The mental model to keep

Document for the hardest reader in the room, not the easiest. A finding written to be reproduced, kept separate from its inference, backed by preserved evidence, and handed identically to everyone is a finding an adversarial expert has to build around rather than knock down. You are not writing to be understood, you are writing to be un-dismissable.

References

• Trade-standard practice for service-record retention on contested and claims-adjacent work
• Manufacturer documentation on diagnostic tolerances and specified conditions
• See related: Documenting a Diagnosis So It Holds Up Against a Warranty Dispute; The Evidence That Settles a Three-Way Technical Dispute