Commission a study that informs a decision, not a quantity of calls or a promise to validate your idea. The brief should define what you need to understand, whose experience is relevant, how those people can be reached and what evidence you will receive. Successful interviews may reveal a sound reason to change or postpone the offer rather than confirm that participants like it.
If enquiries stall after initial contact, “do customers like our service?” is not a useful starting question. Ask what step the prospect was trying to complete, which information or approvals they needed, and where the process stopped. That connects the research to a decision the team can make afterward.
Identify the decision interviews will support
List the options genuinely available: clarify the offer scope, change enquiry-form inputs, reorder qualification, or obtain more evidence. Define what the study will not answer, such as market share or proof that a particular price is optimal.
User Research can organize interviews around these questions and connect findings to product and design decisions. Before committing to delivery, confirm a qualified accountable researcher, legitimate participant access, and suitable consent and incentive arrangements. A service name in a proposal does not establish that recruitment is immediately available.
Agree how negative findings will be handled. If the brief quietly requires a positive conclusion, the work becomes a sales exercise rather than an investigation. Acceptance should concern the quality and traceability of the study, not whether it supports the sponsor’s preferred change.
Define participants through relevant experience
“Business managers” is not a sufficient recruitment specification. For stalled enquiries, participants might need to have made a relevant enquiry within a defined period, including people who continued and people who stopped. Check their role in the decision: an end user may understand the problem without knowing why purchase approval was delayed.
Agree eligibility and exclusion criteria, recruitment channels, invitation ownership, and what happens after a refusal or missed session. Employees or friends should not silently replace target customers. If an important group cannot be reached, show the gap in the report and limit the conclusions accordingly.
GOV.UK recommends recruiting actual or likely users relevant to the service. Participant-recruitment guidance. That is a methodological starting point, not a numerical guarantee of sample quality or market representation.
Worked hypothetical example: enquiries stopping before quotation
A fictional service company wants to understand why some prospects stop before receiving a formal proposal. The proposed study includes different paths: people who supplied the requested information, people who stopped after it was requested, and people who selected another option. Final recruitment depends on the question, case diversity and feasibility, not a universal promised number.
The researcher asks about the last relevant enquiry, what the person was trying to accomplish, whom they needed to consult and what information was requested. They explore the sequence without prescribing the explanation. “Did you stop because the form was long?” plants a cause. “What happened after the request arrived?” leaves room for different accounts.
Suppose one case reveals that the enquirer did not possess the requested site specifications. Record three layers: the participant reported waiting for information from a colleague; requiring all specifications immediately may obstruct some cases; a possible response is to test an initial submission with the information already available. These are a report, an interpretation and a candidate action. They should not collapse into “customers reject long forms.”
If a participant shares a document, establish permission and necessary redaction before using it in analysis. A screen share is not blanket permission to collect their customers’ files or colleagues’ personal information. The purpose is to understand the workflow, not gather everything visible.
A contradictory case matters too. Someone who supplied the same information easily may reveal an important role difference. The study should preserve that distinction instead of averaging both stories into an unsupported general statement.
Treat consent as part of study design
Participants should understand the activity, how data will be used, who can see it, and whether recording or observation is involved. GOV.UK’s informed-consent guidance addresses these elements. Have the responsible specialist review privacy arrangements and local obligations; a foreign methodology reference is not a compliance certificate.
Define the required evidence package: de-identified reporting, restricted notes, or recordings where both permission and need justify them. Do not require every recording to be distributed to the entire team. A traceable summary preserving context may be more appropriate than broad access to sensitive material.
Also agree how Arabic and English evidence will be handled. If a statement is translated, label the translation and preserve its meaning and context. An unclear phrase should be checked rather than rewritten into a stronger commercial claim.
Accept outputs that can support action
Request an approved brief, recruitment criteria, discussion guide, coverage record and findings matrix linked to permitted evidence. Each interpretation should show support, contradictions and unresolved uncertainty. A collection of attractive quotations is not a measure of evidence quality.
If the study uncovers meaningful buyer-segment differences, Market & Offer Research can investigate offer fit more broadly. If it proposes a specific change, Product Validation can test that assumption against an agreed decision rule instead of treating an interview as final proof of demand.
Acceptance checks:
- The report names the questions and decisions it was commissioned to support.
- Participants match explicit criteria and uncovered groups remain visible.
- Reports, interpretations and recommendations are distinct.
- A small qualitative sample is not converted into market percentages.
- Agreed recording, storage and sharing boundaries are documented.
- Recommendations identify a next step or unresolved question rather than an absolute generalization.
Frequently asked questions
Should the interview count be fixed upfront?
An initial range and review mechanism can be agreed, but justify the count through question, case diversity and resources. Count alone does not guarantee saturation or representation.
Can sales staff attend?
Appropriate observation may be arranged with participant agreement, but the session should not become a sales pitch. An active sales relationship can affect what someone feels comfortable saying.
Must every session be recorded?
No. Agree sufficient evidence, workable analysis and participant rights. If recording is unavailable, explain the implications for checking notes.
Can a supplier guarantee validation?
No. You can agree process and output quality, not a preferred answer. Useful research permits negative and inconclusive outcomes.
Start by sharing the pending decision and the customers you can genuinely reach. Agree recruitment and evidence boundaries before scheduling sessions.
