A prelaunch survey should answer one decision, such as which buying situation to investigate next or which delivery window deserves a pilot. Write short, neutral questions, test that people interpret them as intended and document how respondents were recruited. A collection of enthusiastic answers is not automatically a representative market study.
Write the decision before the questions
Suppose you need to choose between weekday collection and weekend appointments. Relevant questions concern recent behavior, scheduling constraints and the person responsible for arranging the service. Asking about brand colors or general interest will not settle that decision.
Also ask whether a survey is suitable. If you do not understand the problem or the language customers use, a few exploratory interviews may help first. If you need to know whether people can operate a prototype, observing a task is more direct than asking whether it looks easy.
Separate experience from prediction
Recent-event questions ask what actually happened within a stated period. Preference questions ask what someone favors. Purchase-intention questions ask what they imagine doing. Keep these categories distinct in both the questionnaire and your interpretation.
For example, “When did you last pay for this task?” differs from “Would you pay for help?” The second may be useful background, but it does not establish a purchase. Avoid writing a forecast that treats every positive intention as a future customer.
Draft a compact question sequence
| Purpose | Possible approach |
|---|---|
| Eligibility | Ask whether the respondent has encountered the relevant task |
| Recent behavior | Ask about the most recent occasion within a clear period |
| Current alternative | Ask how it was handled, with reasonable options |
| Constraint | Ask what limited the available choices |
| Decision input | Ask about the one schedule or service choice under consideration |
| Unexpected detail | Offer a short optional open response |
This is an example sequence, not a universal questionnaire. Include “not applicable” or an appropriate alternative where respondents may lack the experience. Do not force a preference between two options that neither suits.
Remove wording that changes the answer
Avoid double questions such as “Was the service quick and affordable?” A person may agree with one part and disagree with the other. Avoid flattering your own idea with phrases such as “our convenient new solution.” Use concrete periods and familiar terms instead of “regularly,” “soon” or unexplained industry language.
Pew Research Center discusses wording, question order and response choices. AAPOR emphasizes survey design and methodological transparency. Apply those principles carefully, while recognizing that a small founder-run survey will have limitations quite different from a professionally sampled public-opinion study.
Hypothetical example: twenty-four responses
Imagine twenty-four people answer an illustrative survey shared through a neighborhood association. Fifteen report handling the task during the last six months. Nine of those fifteen prefer a weekend appointment. You can report those counts with the recruitment context.
You cannot conclude that most residents will buy a weekend service. The association’s readers may differ from other residents, nonrespondents are unknown and a preference is not a purchase. The result may justify researching weekend delivery with a relevant subset, rather than ordering equipment for a predicted market.
Pretest the questionnaire
Ask a small number of appropriate people to explain what they think each question means. Watch for missing response options, confusing time periods and questions that assume an experience they have not had. Revise before distributing widely, then preserve the version actually used.
If you change wording halfway through collection, keep the versions separate. A response shift may reflect the question rather than the market. Record dates, distribution channels, eligibility criteria and incomplete responses. Avoid collecting names or sensitive details that the research does not need.
Check whether people abandon a particular question. Missing answers can reflect confusion, sensitivity or an irrelevant assumption, rather than simple lack of interest. Keep partial-response counts and review the wording before treating completed questionnaires as the only evidence worth preserving.
Prepare an honest results note
Show the question wording, relevant respondent count, recruitment route and important limitations alongside the finding. Counts are often clearer than percentages when the sample is small. If you show a percentage, include its denominator and do not imply precision the collection method cannot support.
Finish the note with the decision the answers inform and a remaining uncertainty. For example, a preferred time window may still need a test of actual availability and payment. Good survey reporting makes the next question clearer instead of making the business appear more certain than it is.
Survey release checklist
- One practical decision guides the questionnaire.
- Questions distinguish behavior, preference and intention.
- Response options do not force irrelevant choices.
- The wording has been pretested.
- Recruitment and questionnaire versions are documented.
- Results include counts and sampling limitations.
- Follow-up contact requires an appropriate separate permission.
