How to conduct product research: a fast, numbered checklist

Use a tight, repeatable checklist to conduct product research, instead of a long research narrative. This guide gives you six numbered steps, sample question wording, and the survey logic that keeps your data clean.

Work through these steps in order. Do not skip the concept test, since it separates a hunch from a validated decision.

Write down the specific business decision this research supports. "Should we build this feature" is too broad. "Should we add automated meeting recaps to our project management tool" is a decision you can test.

Identify the exact segment whose opinion matters most, such as current users, churned users, or prospects who evaluated a competitor. If you lack a list, source respondents through a targeted panel rather than a convenience sample.

Write a short round of open-ended questions to understand the problem before proposing a solution. This catches needs you have not thought to test, so step four does not rest on internal assumptions alone.

Create a clear, single concept description, mockup, or short demo video for each idea you are weighing. Keep each concept focused on one idea so respondents react to the thing itself, not to a lineup.

Launch your concept to your audience and collect structured, scaled responses alongside a few open-ended follow-ups. Give the survey enough time in the field to reach a sample size that supports a real conclusion.

Review the scored results against your decision from step one, not how the idea felt in the room. A concept testing tool with built-in benchmarks and key driver analysis, like Concept Testing, scores this automatically, so you spend your time on the recommendation instead of the spreadsheet.

Borrow this wording and adjust the bracketed detail for your product.

Pair a few scaled questions like these with one open text field so you get both a number to track and language you can quote back to stakeholders.

Skipping this process is expensive in a way that does not show up until later.

Teams that greenlight ideas on gut feel alone tend to discover the gap only after engineering time and a launch date are spent. A tight, numbered process front-loads that risk. It gives product and marketing leaders a shared answer to "why this idea, not that one," and gives finance a number to cite when a roadmap decision gets questioned.

A SaaS team might test a new automated approval workflow with target customers before a sprint is scheduled, rather than discover after launch that the feature solved a problem nobody had.

Branching keeps a survey short for each respondent while still collecting depth where it matters. A common setup: ask a screener question such as "have you used a project management tool in the last six months." Respondents who answer no skip to a thank-you page, since their feedback would not be reliable for a feature-specific study. Respondents who answer yes branch into the concept description and follow-up questions.

Inside the concept test, monadic testing means each respondent sees only one concept, not all of them in sequence, a form of branching that happens automatically. For directional feedback across multiple rough ideas instead of one locked-in choice, a sequential monadic setup, where each respondent evaluates more than one concept in order, suits that earlier stage better.

For the full stage-by-stage process behind this checklist, see how to perform product research. For a broader listening program beyond a single study, the market research use case overview covers how ongoing programs are structured. To narrow a long list of rough ideas before a full concept test, Idea Screening is built for that earlier stage.

You do not need to build a study from a blank page. Start with a ready-made structure built around the questions that matter most for early product feedback, then adjust the wording to fit your concept. Start from a template to get your survey in front of real respondents today.