In this article

A customer asks for a dashboard. The request could mean several things: they cannot find an overdue order, they need a slide for a meeting, or they are tired of answering the same question from their manager. Each explanation points toward different work. Only some require a dashboard.

The growth opportunity begins to take shape when you can explain the situation behind the request and the consequence of leaving it unresolved. A feature name provides neither. Before estimating the upside of a solution, establish whose behavior might change, when that change could occur, and why the obstacle matters enough to act on.

Reconstruct the last occurrence

Ask the customer to walk through a recent attempt. Where did the need begin? What did they use? What happened when the process failed? A remembered incident gives you something more concrete to investigate than a general statement about what would be helpful.

In a hypothetical order-management product, an operations coordinator might describe exporting a spreadsheet every Friday to find orders awaiting approval. The export itself may be easy. The difficulty could be that the spreadsheet does not identify who needs to approve each order. A dashboard containing the same incomplete data would move the problem onto another screen.

Look for the current workaround and its cost. A coordinator who calls several colleagues has already demonstrated a reason to resolve the uncertainty. That does not prove they would adopt your solution, but it tells you where the effort goes. Ask to see an appropriate example when the research setting permits it; do not turn a confident recollection into a precise time-savings claim.

Strategyzer’s guidance on customer experiments distinguishes accounts of past experience from hypothetical opinions, and verbal evidence from actions. That distinction is useful here. “Would you use an approval dashboard?” invites agreement with your idea. Examining the last delayed order can reveal whether the proposed dashboard addresses anything consequential.

Write the opportunity without naming the feature

A working statement for this example could be: “Coordinators reviewing overdue orders cannot identify the next approver, so they spend the review contacting colleagues and leave some orders unresolved.”

The statement contains a situation and an observable consequence. It leaves room for several responses: better ownership data, a direct request to the approver, an exception report, or a change in the approval process itself.

That openness protects discovery from becoming a justification exercise. If the team has already decided to build a dashboard, every conversation can start to look like dashboard evidence. Put the proposed solution beside the opportunity, not inside its definition.

Also separate customer harm from the commercial hypothesis. Delayed approvals might discourage continued use of the product. They might instead be an operational problem that customers tolerate because the product remains indispensable. The first claim needs evidence beyond the existence of the delay.

A useful growth hypothesis names the nearer behavior first: more overdue orders resolved during the weekly review. Paid retention or expansion can be a later business question. Jumping straight to revenue makes a small product test responsible for proving too much.

Find out who does not have the problem

Repeated stories can make an opportunity feel universal. Deliberately look for comparable customers who complete the task without difficulty. The difference may explain the opportunity better than another complaint does.

Perhaps smaller teams know every approver personally. Perhaps one customer maintains ownership data elsewhere. Perhaps the problem occurs only when an order crosses departments. These circumstances can narrow the target segment and reveal the actual missing ingredient.

Qualitative research can expose that mechanism without establishing prevalence. If several interview participants report a problem, do not convert that share into a population estimate unless the research design supports it. Use appropriately sampled operational data or further research to examine reach.

Watch for the people your recruitment misses. Current power users may have built effective workarounds. Former customers may no longer want to join a session. Record those gaps when deciding how broadly to apply a finding.

Test the uncertainty that could end the proposal

Strategyzer’s value-proposition testing roadmap separates evidence about customer needs, the proposed offer, and willingness to pay. For a team improving an existing product, those distinctions prevent a positive interview from being treated as proof of adoption or commercial demand.

Choose the next test by the unresolved question:

UncertaintyA useful next investigationWhat a positive result would still leave open
Is missing ownership actually causing delays?Trace recent unresolved orders and compare resolved onesWhether your intervention can supply accurate ownership
Can coordinators use the proposed view?Observe realistic tasks in a prototypeWhether it changes real weekly reviews
Does supplying the missing information help?A limited real-use pilot with recorded assistanceWhether the service can deliver the information economically
Will customers pay for the offer?A truthful commercial offer with real termsRetention and cost over subsequent cycles

A manual pilot may be valuable before a large integration project. In the order example, a team could supply an owner list for a defined set of orders with the customer’s agreement. Record where the information came from and how much human effort it required. If the pilot succeeds because a skilled operator privately resolves ambiguous cases, the software proposal still has that problem to solve.

A prototype is appropriate for a different uncertainty. It can reveal whether users understand the owner labels and know what to do next. It cannot establish that ownership data will be correct in production.

Strategyzer’s Testing Business Ideas resource catalogs experiments for different assumptions. Use such a library to widen the available methods, then choose the one that answers the specific question. A familiar test is not automatically the right test.

Make room for a result that changes the plan

Before the pilot, write down what would weaken the opportunity. Accurate owner information might make no difference because approvers lack authority. Customers might resolve the order successfully but reject the time needed to maintain the data. The problem might be severe for a segment too small to support the proposed investment.

Those results should alter the plan. A failed intervention does not necessarily erase the original problem, and a vivid problem does not entitle one solution to endless retries.

The next internal discussion should distinguish what the research has established from what remains open. It might establish that a particular segment loses time identifying owners. A later prototype study might show that the proposed view is understandable, while its effect on real approvals remains untested. Keeping those claims separate gives the team a stronger basis for funding work than a slide headed “Customers want dashboards.”

Sources

Something we should clarify?

Suggest a correction