UX Research on a Budget: 5 Methods That Change Roadmaps
Key takeaway
Five interviews with real users, watched rather than surveyed, will reshape a roadmap more than any volume of analytics. Ask about what people actually did last time, never about what they would do — predicted behaviour is unreliable, remembered behaviour is evidence.
Teams skip research because it sounds like a department and a quarter. The methods that actually change decisions are small enough to run in a week, and the biggest risk isn't sample size — it's asking questions that guarantee agreeable answers.
1. Five user interviews
Talk to five people who have the problem you're solving. Ask them to walk you through the last time they faced it, in detail, including what they used and what annoyed them. Don't describe your solution until the end. Five conversations reliably surface the major themes; the tenth mostly confirms the first five.
2. Usability testing on a prototype
Give five people a realistic task and watch them attempt it without help. Say nothing while they struggle — the silence is where the finding is. This can be run on a clickable prototype before any code exists, which is when changes are essentially free.
3. Support ticket and search log mining
Your support queue and internal search logs are research nobody has read yet. Categorise a few hundred tickets and count. The top three categories are usually a roadmap, and the argument is already backed by volume.
4. Session replay on the failing funnel
Analytics tells you where people drop out; replay tells you why. Watch twenty sessions that abandoned at the step that matters. You'll typically find one concrete, fixable obstacle in the first ten.
5. Fake door and demand tests
Before building a significant feature, put the entry point in the product and measure how many people try to use it — then tell them honestly that it's coming and offer to notify them. This costs an afternoon and repeatedly prevents quarters of misplaced work.
The mistakes that make research useless
- Asking "would you use this?" — people are polite and bad at predicting themselves.
- Talking to people who resemble your team rather than your users.
- Demoing before listening, which converts an interview into a sales call.
- Treating one loud customer's request as a trend.
- Running research after the decision, to justify it.