Every business case that reaches an investment committee is a forecast. Revenue reaches this level by year three, the synergies land by month eighteen, the investment pays back inside six quarters. Then ask the awkward question. When did anyone last go back and score an approved case against what actually happened?
For most, the honest answer is never. Mega-projects get formal reviews and disasters force write-downs, but the ordinary flow of cases that moves most of the capital is argued over for weeks before approval, then never looked at again.
What never scoring breeds
Predictions without feedback drift, and capital allocation shows every symptom. A business case is written to win approval, and its sponsor is rewarded for winning. When nobody scores the outcome, the optimism that wins approvals never gets billed, so each cycle learns that optimism works. The sponsor whose last four cases overshot by 40% walks into the committee with the same standing as the one who delivers. Integration costs get underestimated the same way in every acquisition and ramp-up curves run hot in every launch, because the errors are systematic and unmeasured.
The biggest cost is invisible. Nobody sees the good projects a padded case crowded out.
A committee that never learns which of its forecasters to believe is not allocating capital. It is refereeing storytelling.
Closing the loop
The repair is a discipline, not a reorganisation:
- Capture the forecast at the moment of approval. The revenue curve, the cost line and the milestones become commitments with dates and owners, not paragraphs in a slide deck.
- Track actuals against them automatically. The numbers already live in the company's systems, so scoring runs as a standing pipeline, not a once-a-year project.
- Publish the track record, by case, by sponsor and by type of assumption. The point is not blame. Over many cases promises and results should converge, and publishing the gap is what closes it.
- Feed the record into the next approval. A new case arrives with the sponsor's scored history attached, so the committee debates an evidence-adjusted number instead of a fresh story.
The cases change before the scoring starts
The most valuable effect arrives early. Once sponsors know their cases will be scored, the cases change on the way in. Ranges replace single-point promises, contingencies surface, and the padding drains out because it no longer pays. The committee starts receiving something closer to the truth.
The Prophesee Finance Suite brings this to capital allocation. Commitments are captured in structured form, actuals are tracked against them automatically, and every sponsor's record follows them into the room. Score your last ten cases. Start here.