Most businesses have more provable material than they use and less than they claim. The two conditions coexist comfortably because nobody has ever laid the material out in one place.
A proof matrix does that laying out. Down one axis, every doubt a buyer holds at each stage of the journey. Across the other, every artefact the business could put in front of that doubt: a worked example, a method document, a reference, a sample of the actual deliverable, a published position, a measurement, a credential. The cells get filled with what exists, and the empty cells are the finding.
What the empty cells usually show
The pattern repeats across businesses. Proof is abundant at the stage where the buyer is already convinced and thin at the stage where they hesitate. Case material demonstrates that the work was delivered, which is a doubt the buyer stopped having three stages ago. Nothing addresses whether this firm understands a situation like theirs, which is the doubt that stops the deal.
The second pattern is proof that exists but is unusable, because it was never cleared. A result nobody has permission to name is not proof. It is a thing you know that you cannot say, which is a different asset and often a worthless one commercially.
Building the missing pieces
Once the gaps are visible, most of them are buildable without a client. A method written out in enough detail to be disagreed with demonstrates competence directly. So does a worked example on a hypothetical, clearly labelled as one. Neither is as strong as a named result, and both are available now to a firm that has none it can publish.
The matrix is maintained rather than produced once, because every new engagement changes what can be shown and the change is easy to miss when nobody is keeping the list.