Client-First Constraint Discovery
Uncover the client’s real constraint before optimizing the solution
- Difficulty
- Easy
- Time to result
- ~days to results
- Steps
- 5
- Confidence
- 98%
Begin with the client’s stated objective, but treat it as a hypothesis rather than a complete specification. When a technically strong proposal is rejected, stop defending the model and investigate what the client experiences as failure. Ask direct, nonjudgmental questions about legal rules, incentives, timing, cash flows, internal politics, and operational realities. Translate the answers into a clear binding constraint, then rerun the analysis against that constraint. The output is not necessarily the solution with the highest abstract value; it is the strongest solution the client can actually use. Close the loop by explaining the revised objective and obtaining confirmation before proceeding.
Origin
Heinel learned this while working on an airport authority refinancing. Senior bankers optimized total net-present-value savings, but a short conversation revealed that the authority could retain only first-year savings. Extracted from Masters in Business.
Core principles
- 01The stated problem may not be the real problem.
- 02Expertise does not replace attentive listening.
- 03Operational constraints can outweigh theoretical optimization.
- 04The best feasible answer may differ from the mathematically optimal answer.
How to run it
- 1
Treat the Brief as a Hypothesis
Document what you think the client wants, the metric being optimized, and the assumptions behind it. Do not treat prior client patterns as proof that this client wants the same thing.
- 2
Investigate the Rejection
If the client resists a seemingly strong answer, ask what outcome they expected and why the proposal fails in their environment. Listen for constraints that were absent from the original brief.
- 3
Find the Binding Constraint
Separate preferences from requirements and identify the legal, financial, operational, or timing condition that truly governs the decision.
- 4
Reframe the Objective
Replace the original optimization target with one that reflects the discovered constraint. Confirm the revised problem statement with the client before rebuilding the solution.
- 5
Optimize the Feasible Answer
Design the strongest solution that satisfies the real requirement, even if it scores lower on the original metric. Explain the trade-off explicitly.
In the wild
An airport authority repeatedly rejected a refinancing that maximized discounted savings. A junior banker asked why and discovered that the authority could retain only first-year savings; later savings merely reduced collected fees. The deal was redesigned to front-load the benefit instead of maximizing total net-present-value savings.
→ The transaction aligned its cash-flow profile with the authority’s statutory constraint.
A vendor proposes the most feature-rich deployment, but the client rejects it. Discovery reveals that the binding constraint is a regulatory deadline rather than feature coverage. The vendor narrows the first release to compliant workflows and schedules optional capabilities later.
→ The client meets the deadline with a usable solution rather than missing it with a theoretically superior one.
Common mistakes
Assuming Every Client Wants the Standard Optimum
A familiar metric may not capture the client’s legal, incentive, or cash-flow reality. Validate the objective instead of importing it from previous engagements.
Treating Rejection as Irrationality
Resistance often signals missing information. Curiosity produces better data than dismissing the stakeholder as difficult.
Optimizing Before Confirming Constraints
More analysis cannot rescue a model built around the wrong target. Confirm the problem before refining the answer.
Is it for you?
Best for
Advisers, consultants, managers, and product teams solving complex problems for clients or stakeholders.
Not ideal for
Routine decisions where the objective and constraints are already explicit and independently verified.
From the transcript
“you got to really listen sometimes the problem is not what you thought the problem was uh and sometimes the answer even though it's not…”
“it really helped me if I understood better why this isn't working for you.”
“Nobody asked that question and nobody asked that question right because they just thought they knew better because every other client wanted max net present…”
From the episode
Balancing $5.7T in Active and Passive Management with Lori Heinel