Main idea

An identified hole is information: it can be cited in a meeting, it can be quantified, it sets a reservation on the conclusion — "we know nothing about usage among small accounts, the opportunity is only validated on the mid-market segment". It weighs on the decision instead of distorting it.

An invisible hole does the opposite. Nothing distinguishes it from the rest, so the analysis behaves as if the subject were covered, and the conclusion is presented with the same assurance as a complete one. The error does not bear on what we know, it bears on what we believe we have covered.

This invisibility is not negligence: it is structural. An undocumented area produces no file, no note, no link — by construction it has no manifestation in the corpus. The gap can therefore only be seen if something describes in advance what should have been covered, and compares.

The setup that makes a gap spottable has a simple shape: a counter with three states, not two. An object corresponds to a sheet; an object has been examined and settled as purely technical, decision written down; an object has not been looked at yet. Only the third counts, and it is the one that has to fall to zero — the displayed percentage, for its part, does not have to climb to 100, failing which you create "Queue" and "Token" sheets to push a number up. The sorting between the first two states is what prevents omission by oversight.

Layer added by "I wrote a product's ontology. Three times, I thought I was done." (2026-08-11).

Why it matters

This reverses a reflex: we judge a corpus on what it contains, whereas what decides its reliability is what it declares not to contain.

It also justifies an object usually thought of as bureaucratic — a state of completeness kept alongside the content. It is the only artifact that can make a hole visible, since the hole itself will never manifest.

Nuances and limits

Declaring what should have been covered presupposes being able to draw the subject's perimeter, which is precisely what is hard on a new subject: you do not list the areas you never thought of.

And an inventory of gaps can become a way of never concluding, each empty box serving to push the decision back.

Open questions

• On an exploratory subject, what can stand in for a reference perimeter before you know the subject?