User research

User research is collecting and analyzing data about your real (or potential) users with suitable methods: qualitative (in-depth interviews, observation, usability observations) and quantitative (surveys, behavior analytics), for specific questions. The result is an understanding of the audience, its tasks, pains, context, barriers, formed into conclusions for product decisions. Honestly about 'informs, does not predict', this is key: research reduces uncertainty and grounds decisions but does NOT predict the future with 100% accuracy or guarantee the product will 'take off'. It says what exists NOW for the studied people, helping make bets more deliberately rather than blindly — but it is probabilistic knowledge, not a guarantee. Honestly about sample and methodology: the quality of conclusions directly depends on who and how was studied. A small or unrepresentative sample gives limited conclusions (you cannot boldly generalize to everyone); poorly framed questions give distorted answers. We pick the method and sample for the question and honestly flag the limits of conclusions' applicability rather than pass off the opinion of 5 people as 'the opinion of all users'. Honestly about 'say ≠ do': people do not always act as they declare (rationalization, social desirability). So observing behavior matters, not just asking — we account for this rather than take words literally. Honestly about the effect: it gives a grounded picture for decisions and reduces the risk of making the unneeded, but it is an input for decisions, not a product-success guarantee. Honestly about access: access to users/the audience and an understanding of the research questions are needed. An important boundary: this is research; personas/JTBD/CJM — artifacts based on it (944/945/946); usability testing of the ready product — 953. Picture this: instead of 'deciding for users what they need' — understanding based on their real behavior and words. The base price starts from 45,000 ₽ (depends on methods and scope).

• Product decisions are made on guesses about users.
• It is unclear what the audience really needs and where its pains are.
• What seems needed is done, not what people need.
• No data, only opinions inside the team.

• Choosing methods for the questions (interviews/observation/surveys/analytics)
• Collecting data from real/potential users
• Analysis of tasks, pains, barriers, context
• Conclusions for product decisions with applicability limits
• Accounting for 'say ≠ do' (observation, not just words)
• Honest boundaries (informs, does not predict; depends on the sample)
• A link with personas/JTBD/CJM (944/945/946)
• Handover and review with you

• A grounded understanding of the audience and its tasks
• Decisions on facts, not on guesses
• Reduced risk of making the unneeded
• Honest boundaries (an input for decisions; not a success guarantee)

• We formulate research questions; pick methods and sample
• We collect and analyze data (words + behavior)
• We form conclusions with limits, honestly set constraints with you

• Will research tell exactly what to do to make the product take off?

  No, honestly: research informs decisions but does not predict the future or guarantee success. It shows what exists now for the studied people, helping make bets more deliberately rather than blindly — but it is probabilistic knowledge. We reduce uncertainty and ground decisions, while guaranteeing 'it will take off' would be deception — success depends on execution and the market.

• Is interviewing 5–10 people enough to know about everyone?

  Depends on what for, honestly. Qualitative research (interviews/observation) on a small sample excellently uncovers problems and patterns but does NOT give statistics of 'what % of users'. For generalizations to everyone quantitative methods and a representative sample are needed. We honestly pick the method for the question and flag what can be generalized and what cannot — rather than pass off a few people's opinion as 'everyone's opinion'.

• Can we just ask users what they need?

  Asking is useful, but honestly: people do not always do what they say (rationalization, social desirability) and do not always realize their real needs. So it is important not only to listen but to observe behavior. We combine methods and do not take words literally — that way conclusions are more reliable than 'users said so, therefore it is so'.