Separate Buyer Problems from Assumed Root Causes
Use a symptom-cause worksheet to test competing explanations, with a worked late-report example that redirects discovery from software speed to definitions.
10/23/20263 min read
A buyer says that reporting takes too long. That statement describes a problem they experience, but it does not identify its cause. The delay could come from collecting data, correcting definitions, waiting for approval or deciding what the report should contain. Selling a faster reporting tool before separating those possibilities risks solving the wrong part of the process.
A symptom-cause worksheet keeps the buyer's observation intact while making the seller's explanations testable. It should produce a better diagnostic question, not a premature diagnosis or a claim that the proposed service must be needed.
Preserve the observation first
Write what happened in the buyer's terms, including the relevant process and time period. Keep interpretation in a separate field. Monthly reporting was late twice is different from our database is too slow. If the buyer uses a causal explanation, record it as their current interpretation unless evidence has established it.
Ask for one recent example that can be discussed without unnecessary personal or confidential information. Trace where the work entered the process, where it waited and what happened next. Do not collect a full data export when a sequence of steps and approximate timings would answer the immediate question.
Build competing explanations
List a small number of plausible causes that would lead to different interventions. For each, identify evidence that would support it and evidence that would weaken it. Avoid a list of explanations that all conveniently require buying your service. Include the possibility that a simple process clarification or no intervention is appropriate.
A useful worksheet has five columns: observed symptom, possible cause, distinguishing question, available evidence and current status. Status can be supported, weakened or unresolved. It is not a probability score unless the team has a justified method for producing one.
Worked example: the late report
In a fictional discussion, Stone Process hears that a buyer's monthly report takes five working days to finalize. The initial seller hypothesis is slow extraction. The buyer's walkthrough reveals that extraction finishes on the first morning, while two departments spend several days reconciling different definitions of active customer.
The completed worksheet contains three hypotheses. Extraction delay: weakened by the same-day export record. Definition mismatch: supported by conflicting department definitions and repeated reconciliation. Approval availability: unresolved because the approval timestamps were not reviewed. The next useful question concerns who can agree the shared definition, not which new extraction tool to demonstrate.
The seller proposes a bounded review of the definition and one synthetic reconciliation example. No promise is made that this will eliminate every delay. If approval availability later proves important, it becomes a separate workstream rather than evidence that the original guess was correct all along.
Avoid confirmation by repetition
A claim repeated in a meeting, CRM note and proposal may still come from one untested assumption. Keep the source of each observation visible. W3C's provenance framework provides a useful conceptual reference for preserving how information was produced; the worksheet here is a proposed practical application, not a prescribed standard.
Ask what finding would change your recommendation. If no possible answer would alter the proposed solution, the exercise is functioning as a sales script rather than diagnosis. Let the buyer correct the map and reject interpretations that do not fit their experience.
End with the next discriminating question
Do not try to resolve every cause in one call. Choose the smallest evidence step that separates the most consequential alternatives. Identify who can provide that evidence and whether doing so is worth the effort. Leave missing facts visibly unresolved.
The finished artifact is a symptom preserved accurately, several competing explanations and a proportionate next check. It helps both sides decide whether deeper work is justified. Sometimes the best result is discovering that the seller's service is unnecessary; that is a legitimate conclusion of a useful diagnostic conversation.