How we grade a finding

Most PMO assessments hand you a maturity score. We grade claims. A dramatic number from a thin sample is not the same thing as a confirmed problem — and we will say so in the report.

Confirmed, Partial, Rejected, or Unfalsifiable. We will not present a thin sample as a settled fact.

Four verdicts

Confirmed — The claim is supported by direct or strongly corroborated evidence. Enough data, and a defined threshold was actually crossed.

Partial — Some evidence supports the claim, but the extent or the cause is still uncertain. Treat as a signal, not a verdict you bet the portfolio on.

Rejected — The evidence does not support the claim, or reliable evidence contradicts it. Rejected is not "low severity." It means this specific concern is not substantiated.

Unfalsifiable — There is not enough information to test the claim. Unfalsifiable is not true and not false. It means you are steering without a measurement. That is often the most expensive finding in the room.

What we will not do

  • ●      Blend a measured dollar figure with a judgment call into one impressive ROI.

    ●      Let an AI invent a number that is not in your extract.

    ●      Present a one- or two-project sample as Confirmed.

    ●      Recommend only additions. Every finding names something to stop.

The catalog of breakdown patterns and the interview stay with the engagement. The certainty math is public because you should be able to check us.