Present an AI investment recommendation with uncertainty

An investment recommendation is credible when leaders can see which assumptions drive it and what evidence would reverse it. A single precise return number hides the uncertainty that a pilot is meant to reduce.

Build the decision artifact

Show the baseline and no-change option. Build low, base and high cases for eligible volume, adoption, quality, exception cost, platform cost and support. Separate reversible pilot spend from scale commitment. Identify the variables with the largest sensitivity and attach gates to them.

The commercial leadership lab makes the artifact inspectable with authored inputs:

from leadership_cases import investment_case

result = investment_case()
assert result["net"] == {"low": -40, "base": 120, "high": 250}
assert result["recommendation"] == "bounded pilot with adoption gate"
assert result["guaranteed_return"] is False
assert result["currency"] == "fictional thousands"

The fictional net result ranges from minus 40 to plus 250 thousand units. Rather than asserting the base case, the fixture recommends a bounded pilot with an adoption gate. It makes no promised return.

Protect the decision from weak evidence

Scenario labels can disguise arbitrary optimism. Ground ranges in observed workflow data or explicit experiments, not labels alone. Include opportunity cost, delay and exit; show the decision if the low case arrives.

Keep these artifacts for review:

• baseline and no-change comparison
• scenario assumptions and sensitivity
• pilot cost and learning objectives
• recommendation plus reversal conditions

This practice aligns with the discovery, productisation, client enablement, technical leadership and capstone sequence in the FDE for Professionals course. The course link describes the pathway; the local scenario is fictional and does not claim a client engagement, investment result, hiring decision or certificate.

Practice task

Change adoption in the base case by ten points either way. Explain whether the recommendation changes and which measurement should be collected first.

Continue learning

This article is part of the Commercial judgement and delivery leadership sequence. Use the neighbouring tasks when you need the prerequisite or the next application.

• Review the prerequisite or neighbouring task in Design a technical due-diligence checklist for an AI pilot.
• Continue with Explain a failed pilot without hiding the original assumptions.

Reference: GOV.UK Service Manual: Measuring Success.