Skip to content
RD-PREP Tell me about your decision
The RD‑PREP method

Measure to decide, not to reassure.

Uncertainty never disappears entirely. The RD‑PREP method determines what sufficiently reliable information is needed to justify the next decision. The verification effort stays proportionate to what is at stake.

Three questions before committing more resources

  1. What decision really needs to be made?

    The announced decision can hide a more immediate one. Should we develop the product, build a prototype or first test an assumption? Clarifying this step avoids measuring the wrong thing.

  2. What information could actually change that decision?

    The verification focuses on the assumptions that could lead you to proceed, modify, pause or stop. The result that would change your mind is defined before the verification.

  3. What is the simplest way to obtain that information reliably enough for the decision?

    An interview, a calculation, a targeted test or a specialist consultation may be enough. A more complete study becomes relevant when the stakes and the level of evidence sought justify it.

A reversible, low-cost decision does not require the same proof as a major commitment.

Sort the information before looking for it

Two questions help target the verification: could this information change the decision? Can it be obtained reasonably, given the cost, the time and the reliability sought?

Would change the decision, can be obtained

To verify

This is a priority for the next verification. The expected result and the decision criterion are specified before committing resources.

Would change the decision, cannot be obtained

Residual uncertainty

It is named, along with its possible consequences. You must then decide whether the risk can be accepted, whether the commitment should be reduced or whether the decision should be paused.

Would not change the decision, can be obtained

Available, but irrelevant to this decision

Its availability alone does not justify an expense. It may serve a later step without being needed for the current decision.

Would not change the decision, cannot be obtained

Set aside for this verification

No further effort is justified for this decision. The question can be reviewed if the context or the decision changes.

A defensible decision also states what will remain unknown at the time of deciding and the risks this implies.

Separate the result from what it allows you to conclude

The result of a verification is presented separately from its interpretation. Each result is tied to the uncertainty it can actually reduce.

  1. The result

    What was observed

    Observations, measurements or answers are presented with the conditions under which they were obtained and their limits. An assumption remains an assumption until the verification supports it.

  2. The interpretation

    What the result supports

    The result is compared with the criteria set in advance. The analysis states what it supports, what it calls into question and what it does not allow you to conclude.

  3. The decision

    What can be justified next

    The recommendation takes into account the level of evidence reached, the constraints and the residual uncertainty. You then decide which resource to commit and which risks to accept.

Evidence counts only for the decision it supports. A working prototype can demonstrate technical feasibility without demonstrating that a market exists.

  • Proceed
  • Modify
  • Pause
  • Stop

A targeted method, with explicit limits

A scope tied to the next decision

The decision diagnostic examines what is needed to justify the next step. It is not a business plan or an exhaustive study of the project. Questions left open are identified in the brief.

A defined place for specialized expertise

When a question requires specific expertise, RD‑PREP defines the question to be resolved and the level of verification sought. The diagnostic thus helps target the right specialist's involvement.

AI-generated leads must be verified

AI can help formulate hypotheses, organize information or suggest leads. A generated answer is not, on its own, evidence. The elements that determine the decision must be verified.

A conclusion that can challenge the project

Criteria are set so that unfavourable results are recognized too. The verification can support proceeding, lead to a modification, justify a pause or recommend stopping.

What decision do you need to prepare?

Describe the decision to be made, the next resource to commit and what you already know. I will tell you whether a decision diagnostic can be useful.

Tell me about your decision