Find the question behind the request
In supply chain, operations, and transformation work, I translate business questions into requirements that technical teams can build from. That means understanding the decision, testing the definitions, and protecting the business meaning as the request becomes a design.
A useful starting question is: “What will you be able to decide once you can see this?” Stay with the answer until it names a choice someone is responsible for making. Then ask what happens today when that person cannot make the choice confidently.
Build a brief both sides can challenge
Before discussing a dashboard layout, prepare a short reporting brief. Use ordinary language and leave visible space for unresolved questions.
- The user and decision. Identify who will use the information and the decision they need to make.
- The current workaround. Describe how that decision is made today, including conversations, manual checks, and assumptions.
- The required detail. Agree what the user needs to compare, distinguish, or investigate. Include the level of detail that would be excessive.
- The evidence available. Identify which information exists, who understands it, and what still needs verification.
- The acceptance exercise. Describe how the team will demonstrate that the result supports the intended decision.
The brief should be short enough to discuss together. Its value comes from making disagreement visible while it is still inexpensive to address.
Illustrative example: a request for workload visibility
Illustrative example: A manager asks for a workload dashboard to decide whether requests should be reassigned. Counting open requests may help, but the reassignment decision also depends on readiness, specialist skills, and upcoming commitments.
The team could begin with a small view that distinguishes work ready to proceed from work waiting on another party. It could then identify which ready items require a particular skill. This gives the manager a more precise reassignment conversation.
That is a proposed design to test. The team still needs to establish whether the necessary information is maintained reliably and whether reassignment is within the manager’s authority.
Prototype the decision before the screen
Use a simple table and a few representative records. The first prototype can be deliberately plain. Ask the intended user to work through an actual decision using it.
- What do they examine first?
- Which detail changes their interpretation?
- Where do they need to leave the view and ask someone a question?
- What action can they take after reviewing it?
Observe the reasoning as well as the requested changes. A request for another filter may be a clue that two different decisions have been combined in one view. Separate those needs before building a more elaborate interface.
Record the tradeoffs
Every design has boundaries. A view may be suitable for today’s assignment decisions while being insufficient for historical comparisons. A category may require a person’s judgment until the process captures a more precise input.
Document those limits beside the agreed use. Explain what the information supports, which questions remain open, and how uncertainty should be handled. Avoid letting a convenient prototype become an authoritative answer to questions it was never designed to address.
Keep a small decision record: the issue, the options considered, the choice made, the owner, and the reason. That record gives future changes somewhere solid to begin.
Give meaning an owner
As the reporting moves into use, agree who resolves a disputed definition and who communicates changes. The analyst can explain the calculation. The operational owner can confirm what the measure needs to represent. The accountable decision-maker can confirm whether it remains useful.
Schedule a review after people have used the result in real work. Ask what decisions became easier and where they still needed an explanation outside the report. Those answers are the next design brief.

