Organizing sample photos and identification records

A sample valve arrives for evaluation. It gets tested, it works, it goes in a drawer. Eighteen months later someone needs fifty of them and the only record is "the sample from the Chinese supplier".

The fix is twenty minutes on the day it arrives, and it is the same twenty minutes every time.

What to record on day one

Photograph it before anything else - Whole valve from two sides, both connections close up, and every marking. Good light, plain background, no hands in frame.

Transcribe every marking in two columns - What it says, character for character, in one column; what you think it means in the other. Never edit the first column to agree with the second.

Record where each marking sits - Body, bonnet, nut. The same string in two places can mean two different things.

Record where it came from - Supplier, contact, date, reference number, what you asked for when you asked for it. This is how you find the thread again.

Record what was confirmed in writing - The model designation the supplier stated, and the specification they confirmed. Separately from what you measured yourself.

Give it a shelf location - A sample that cannot be found is a sample you do not have.

Keep measured and stated apart

Keep measurements you took yourself clearly separate from values the supplier stated. Both are useful and they are different kinds of fact: yours describes this item, theirs describes the model. When they disagree that is worth investigating, and you can only see the disagreement if they were recorded apart.

Never write a measurement you took into the field where the supplier's stated value belongs. That is how an approximate figure becomes a specification.

The whole thing is half a page

One file, six fields, one folder of photographs. It fits on half a page and it turns a future reorder into a single email with a designation in it.