To Survey or Not to Survey: That is the Question

Surveys are the default research instrument for most teams, which is exactly why so much customer research goes unused. A survey measures how many. It rarely explains why.

When a survey is the right instrument

• You already know the possible answers and need to size them
• You need a tracked metric over time, such as satisfaction or repeat intent
• The population is large enough for the result to mean something
• The decision waiting on the data is genuinely quantitative

When to interview instead

• You do not yet know what the options are
• You are exploring language, motivation, or hesitation
• The audience is small and high-value
• The last survey produced clean numbers and no clarity

Five customer conversations usually reshape positioning more than five hundred survey responses. Conversations produce sentences you can put on a page.

Writing a survey worth sending

Ask fewer questions than you want. Seven or fewer, one screen, no matrix grids.

Ask about behavior, not intention. "What did you do last time" beats "what would you do."

Leave one open field. The verbatim answers frequently carry the real finding.

Avoid leading scales. If every option is positive, your data is decoration.

State how long it takes, and be accurate. Overstated length costs completion; understated length costs trust.

Closing the loop

Send results back to respondents with one sentence on what changes because of them. Customers answer the next survey when the last one visibly mattered.

The question is not survey or interview. The question is what decision waits on the answer, and which instrument produces language you can act on.