Market research survey questions that count what buyers do, not what they promise

Market research survey questions for sizing, pricing and positioning, with the wording to use, why each works and how to check the answers before you act.

Instant Expert EditorialPublished Updated 6 min read

The best market research survey questions count what buyers already do, not what they promise to do. Pick them by the decision they inform: sizing questions count facts respondents know, pricing questions separate what people pay now from how they react to your price, and positioning questions find what people compare you with. Decide in advance how you will check each answer.

For planning, sampling and sending the survey itself, see market research survey.

The example: software for rented equipment

Here is a made-up example used throughout. You are considering software that helps mid-size general contractors track rented equipment such as lifts, generators and excavators. You have heard that rental bills often include days when a generator sat on a site doing nothing but looking expensive. You plan to survey project managers and finance staff at contractors.

1. Sizing questions should count rates you can't look up

A survey can't tell you how many contractors exist. Count those from public data (market size research shows how) and use the survey for the rates nobody publishes: what share rents equipment, how often, and how many have the problem.

  • Qualifier: "Does your company rent heavy equipment for its projects?" Yes / No / Don't know
  • Frequency: "In the last 12 months, about how many separate equipment rentals did your company have?" Ranges such as 1 to 10, 11 to 50, 51 to 200, more than 200, don't know
  • Spend: "Roughly how much did your company spend on equipment rentals in the last 12 months?" Ranges, plus "Don't know"
  • Current process: "How does your company currently keep track of which rented equipment is on which site?" A list built from interviews, plus "Other (please specify)"
  • Problem incidence: "In the last three months, did your company pay for rental days when the equipment was not being used?" Yes / No / Don't know

Then check them:

  • Route each question to the person who knows. If most project managers answer "don't know" on spend, ask finance staff instead.
  • Look for answers that contradict each other. Five rentals and a huge annual rental bill may mean a misread question, or a much bigger company than you screened for.
  • Keep ranges honest. AAPOR recommends options that don't overlap and cover every reasonable answer, including "don't know" where people may need it.

2. Pricing questions should start with what people pay today

Ask about real spend first: "What does your company currently pay for tracking rented equipment, including any staff time or software?"

If the product is new enough that nobody knows what it should cost, the Van Westendorp price sensitivity meter is a common starting point. Sawtooth Software's guide lists its four questions:

  1. At what price would you consider the product so inexpensive that you would feel the quality couldn't be very good?
  2. At what price would you consider it a bargain?
  3. At what price would you say it is starting to get expensive, but you still might consider it?
  4. At what price would you consider it so expensive that you would not consider buying it?

Sawtooth is candid about the limits. The method suits products new to the world, while monadic price tests and conjoint analysis are usually better for established products. It also notes the questions lack competitive context, and that the traditional line-crossing analysis has been criticized as lacking good theory.

3. Discount every "yes, I'd buy it"

Sawtooth notes that respondents typically exaggerate purchase intent, so researchers discount the answers. One possible set of discounts it gives: count 75% of "definitely would buy," 25% of "probably would buy" and 10% of "might or might not buy."

In a made-up survey of 200 contractors where 30 say definitely, 60 say probably and 50 say might or might not:

  • 0.75 × 30 = 22.5
  • 0.25 × 60 = 15
  • 0.10 × 50 = 5
  • Total: 42.5 likely buyers, or about 21% of 200

The raw "definitely or probably" figure would have been 90 of 200, or 45%. The factors are one example, not a rule, but the gap shows why a survey "yes" shouldn't go straight into a revenue model. Researching willingness to pay covers testing price through real actions.

4. Positioning questions should ask open questions before showing any list

  • Current alternative: "The last time rented equipment sat idle on one of your sites, what did you do about it?" Open text
  • Tools considered: "Which tools or services, if any, has your company used or considered for tracking rented equipment?" Open text, before showing any list of names
  • What matters most: Show two criteria at a time, such as "fewer idle rental days" versus "less time spent reconciling invoices," and ask which matters more. Rate every criterion on its own and everything looks important.
  • Problem framing: Show two or three short descriptions of the problem in random order and ask, "Which of these best describes a problem your company has?"
  • Their words: "How would you describe this problem to a colleague?" Open text

Order matters. Pew Research Center found that when closed questions on a topic come before an open one, people are much more likely to mention ideas from the earlier questions. Pew also notes that people filling in surveys alone tend to pick options near the top, which is why it randomizes options with no natural sequence.

The list itself shapes the result. In a Pew experiment, 58% chose the economy as their top issue when it was on a list, but only 35% named it without a list, and 43% of the open-ended answers weren't on the closed list at all. A competitor list you wrote yourself will miss what buyers actually use, which is often a spreadsheet, a rental company's own portal or nothing. Competitor analysis covers researching those alternatives.

5. Check the answers before you act on them

  • Pretest the questionnaire. AAPOR recommends cognitive interviews with people like your respondents, who explain how they understood each question and arrived at an answer.
  • Say where the sample came from. A survey sent to a trade association list or an online panel describes those respondents. AAPOR notes nonprobability samples need special care in analysis and reporting.
  • Compare only groups big enough to compare. Splitting results by company size needs enough responses in each group. Quantitative market research covers sample planning.
  • Call the surprising respondents. Pick five to eight people whose answers you didn't expect and ask about a recent, specific event. If nobody on the phone can name a recent idle rental day, trust the calls and revisit the wording.

Your next step

Write down the decision your survey supports, then pick two or three questions from the matching section. Build answer lists from a few interviews first.

If you need people who aren't on your own list, such as controllers or equipment managers at contractors, describe them to Instant Expert; for surprising respondents you already have, paste their emails. Two steps above fit a paid call: the cognitive pretest, where someone talks you through how they read each question, and the follow-up with surprising respondents. Book a 30-minute call for each. You set what each person receives, our fee is added on top, and you're charged only when someone books. People who accept a paid call are a sample of their own, so keep what you hear on the calls out of your survey counts. The directory page for finance professionals in construction is one place to start.