If You Can’t Explain Where a Number Came From

Here is a question that quietly decides how far an organisation can go with its data: if you cannot explain where a number came from, should an executive make a decision with it. Most people, asked directly, will say no. And yet decisions are made every day on figures whose origins no one in the room could actually trace.

Lineage — the ability to follow a number back through every transformation to its source — sounds like a technical concern, the sort of thing that lives in a tool and interests only engineers. It is in fact one of the most senior concerns there is, because it is the difference between a number you can defend and a number you merely hope is right.

The board-pack number no one can trace

Most people who have spent time near executive reporting have seen the same scene. A figure appears in a board pack. Someone senior asks a reasonable question about it — why it moved, what it includes, how it was derived. And the room goes quiet, because the honest answer is that the number passed through several hands and systems and no one present can reconstruct the path. The figure may well be correct. But no one can prove it, and in that moment its authority evaporates.

Lineage is what prevents that silence. When you can show where a number originated, what was done to it, and who is accountable for each step, a difficult question has a calm answer instead of an awkward pause. That is not a documentation nicety. It is what allows the number to carry weight, and it is inseparable from data quality: you cannot vouch for the quality of something whose journey you cannot describe.

The case becomes sharper still with AI in the picture. When a model consumes data from dozens of sources and produces an output at scale, the question of provenance does not disappear — it multiplies. If you could not trace a single number through a reporting chain, you certainly cannot trace a model's reasoning through an opaque pipeline. Lineage is the groundwork that makes AI explainable rather than merely confident.

None of this requires treating every field in the organisation as sacred. It requires deciding which data matters enough that a decision might rest on it, and making sure that data can be traced end to end. The test is simple, and worth applying more often than it is: if someone senior asked where this came from, could we answer without guessing. Where the answer is yes, trust follows. Where it is no, that is the work — and it is worth doing before, not after, the number ends up in front of someone who has to act on it.