User Research for Founders: Build What People Want

Building a new product can feel like working with the lights off. You have an idea you care about, but one question keeps coming back: “Will anyone actually use this?”

That’s what user research is for. It helps you replace guesses with real signals from the people you want to serve, so you can make better product decisions faster.

Everything starts with one thing: the right question. Not twenty questions. Not a broad “What do you want?” but one focused question that forces you to learn what matters right now.

Defining your one critical research question

You’re ready to talk to customers. That’s a big step. Many founders either delay it too long or show up with a scattered list of questions and hope something useful appears.

Both paths usually end the same way. You get a pile of notes, mixed opinions, and no clear next move.

Before you recruit anyone or write a script, pause and find your “one thing.” What is the single unknown that would give you clarity if you answered it this week?

This is not about solving everything at once. It is about finding the one domino that changes the rest of your decisions.

Here’s a real example. A founder was convinced he needed to build a new dashboard feature. Users were “asking for it,” and it sounded like the obvious next step.

His first question was broad: “What new features do my users want?”

Founder tip: A vague question like “What do users want?” usually produces vague answers. Aim to uncover the underlying problem, not a feature wish list.

We reframed his focus into something specific: “What is preventing new users from completing onboarding successfully?”

It felt like a detour. It turned out to be the fastest path to truth.

Within a few sessions, the answer was clear. People were not asking for a better dashboard. They were getting lost in a confusing signup flow.

The “dashboard request” was a symptom. The root cause was onboarding friction.

That is the goal of this step. You move from a tangled set of ideas to one clear target. Then every conversation stays sharp and useful.

How to frame your own question

Start with your biggest current risk. What belief, if wrong, would hurt your product the most?

Use one of these prompts to get specific:

• “We assume our users are trying to achieve…”
• “We believe our biggest usability issue is…”
• “We are not sure if people understand the value of…”

This forces you to name the uncertainty that is driving your roadmap.

In the dashboard example, the hidden assumption was, “People sign up successfully, they just want more data.” That was wrong. Catching it early saved months of building the wrong thing.

Building a product is a series of decisions. Research helps you lower the risk of each one.

Choosing the right method for your goal

Once you have your one question, the next step is picking the best way to answer it.

There are many research methods, and you do not need to master them all. As a founder, you need a small set you can run quickly and repeat often.

Think of methods like tools. The right tool depends on the question. If you pick the wrong one, you may get clean data that answers the wrong thing.

We see this often in our work. Teams move faster when they match the method to the decision, then ship with confidence.

The four most practical methods

For most of the 100+ products we’ve built, founders get most of what they need from four methods.

• Customer interviews: One-on-one conversations that uncover the “why” behind behavior. Use these to learn motivations, frustrations, and context.

• Usability tests: You watch someone try to complete key tasks using your product or prototype. Use these to find where people get stuck or confused.

• Surveys: Good for measuring “how many” and “how much.” Surveys work best after interviews, once you have a clear hypothesis to validate.

• Diary studies: Participants log experiences over several days or weeks. This helps you understand habits and real-life context you can’t see in a single session.

Better results often come from mixing methods. Interviews help explain the “why,” and surveys help measure scale. Together, they reduce the chance you ship features people ignore.

Which method should you use?

This table helps you match your question to the right approach.

Method	Best for answering	When to use it	Real-world example
Customer interviews	“Why?” “Tell me about…” “What motivates you to…?”	Early, when you need to understand context, needs, and feelings.	“Why do you shop for groceries online instead of in-store?”
Usability tests	“Can people use this?” “Where do they get stuck?”	When you have a prototype or live product and need to spot friction.	“Show me how you’d add three items and check out.”
Surveys	“How many?” “How important?” “Which option?”	After qualitative work, when you want to validate patterns at scale.	“On a scale of 1–5, how important is free shipping?”
Diary studies	“How does behavior change over time?” “What’s the daily context?”	When the journey spans days or sessions and context matters.	“Log every food delivery order for one week and why you chose that service.”

Picking a method is not about being “advanced.” It is about answering your most urgent question with the least waste.

A real-world example: cart abandonment

Imagine you run an e-commerce site. Your analytics show people dropping off at the final checkout step.

Your key question is simple: “Why are people abandoning their carts right before paying?”

Start with interviews to get the human context. Someone might say, “I was excited, but the shipping and fees felt like a bait-and-switch.” That’s emotion and meaning you will not get from funnels alone.

Then run a usability test to see exactly where the flow breaks. You might notice most people hesitate on the shipping options page because the choices feel unclear.

Finally, use a survey for scale. Send it to people who abandoned carts and measure how often fees, shipping, trust, or payment options show up as the top reason.

Used together, these methods turn a vague problem into a clear plan.

Finding and recruiting the right participants

You can have a perfect question and the right method, and still fail if you talk to the wrong people.

This is a common founder mistake. In a rush, it is tempting to recruit friends, family, or anyone who responds.

Those people often know too much about your idea. Or they may avoid giving tough feedback because they want you to succeed.

Your goal is to talk to people who look like your future customers. Their feedback is grounded in real constraints, not politeness.

Where to find your first participants

You do not need a big budget. Start with the pools closest to you.

• Your existing audience: An email list, early sign-ups, or even a small social following can be enough. These people already raised their hand.

• “Watering holes”: Look for niche subreddits, LinkedIn groups, Slack communities, and forums. Join conversations first, then ask for help.

• Second-degree connections: Ask friends and peers for warm intros. A trusted intro often beats cold outreach.

This approach is fast, low cost, and repeatable.

Crafting a screener to filter for quality

Once you have a pool of possible participants, filter them with a screener survey.

A good screener is not just demographics. It helps you find the right behaviors and attitudes, so you learn from people who truly match your target user.

1. Start broad, then narrow: Ask about role, workflow, or habits before asking about your specific problem. This reduces “leading the witness.”

2. Ask about past behavior: Replace “Would you use X?” with “Tell me about the last time you solved Y.” Past behavior is more reliable than opinions.

3. Add one open-ended question: Ask something like, “What is the most frustrating part of doing X?” You will quickly spot who can explain their thinking.

A common trap is recruiting people who are eager to praise. Slightly skeptical participants often give the most useful feedback.

What to offer as an incentive

You are asking for someone’s time. Pay them fairly.

For a 30–60 minute interview or test, $25–$75 as a gift card is common. For harder-to-reach roles, you may need to offer more.

If money is tight, offer a free plan, a discount, or a donation to a charity they choose. The goal is simple: show respect for their time.

This practice also supports long-term retention work, especially when you pair it with a thoughtful subscriber onboarding journey.

Running sessions that actually work

You’ve set your question, picked a method, and found the right people. Now it is time for the conversation.

This is where many founders feel nervous. That’s normal.

The session works best when it feels like a focused coffee chat, not an interrogation. Your job is to make the participant comfortable enough to share the truth.

You want the good, the bad, and the confusing parts. Those details are what improve the product.

The anatomy of a strong 45-minute session

A good session has a clear flow. That keeps you on track and helps the participant relax.

1. Welcome and rapport (5–7 minutes): Thank them, explain the purpose in plain language, and remind them there are no right answers. Ask for consent to record.

2. Context first (10–15 minutes): Ask about their work, habits, and what they do today. Understand their world before you talk about your solution.

3. Core tasks and questions (15–20 minutes): This is where you test the prototype or dig into the main topic. Focus on real experiences and specific moments.

4. Wrap-up (5 minutes): Ask for final thoughts, explain next steps, and send the incentive. End with appreciation.

The most common mistake is talking too much. Aim to speak about 20% of the time and listen for the other 80%.

Questions that get stories, not one-word answers

Your insights will only be as good as your questions.

Avoid leading questions like “Is this cool?” Also avoid future prediction questions like “Would you use this?” People are not great at predicting their future behavior.

Instead, ask about real events. Stories contain context, emotion, and workarounds.

• “Tell me about the last time you tried to [task]…”
• “Walk me through how you handle [problem] today.”
• “What was the most frustrating part?”
• “Can you show me how you’d do that right now?”
• “What surprised you about that process?”

These prompts help people recall real details. Those details lead to better product decisions than simple opinions.

Moderation: gentle guidance

Moderation is a skill. You will get better every time you run a session.

Stay neutral in your tone and body language. Encourage with simple cues like “Tell me more” and “What happened next?”

Silence is also useful. If someone pauses, give them a few seconds. Many of the best insights show up right after a pause.

Turning messy notes into actionable product decisions

Now you have recordings, quotes, and pages of notes. This is where many teams get stuck.

The goal is not a long report that no one reads. The goal is clear patterns and a short list of decisions you can act on.

You want to move from “interesting feedback” to “here’s what we will change next.”

From chaos to clarity: a simple synthesis method

You do not need fancy software. A spreadsheet can work. A whiteboard tool like Miro can make it easier to collaborate.

Here’s a simple process that works for most small teams:

• Extract observations: For each participant, pull out key quotes, pain points, and behaviors. Put each one on its own “sticky note” entry.

• Group similar notes: Move notes into clusters. Look for repeats and patterns. Common groups include “Signup confusion,” “Pricing doubts,” or “Workarounds in spreadsheets.”

• Name each theme: Give each cluster a clear label. Naming turns a pile of comments into a set of findings.

This is simple pattern-finding. It keeps you honest and reduces the chance you overreact to one loud opinion.

Structured synthesis lowers risk. It helps teams validate ideas before committing weeks of build time, which reduces rework and keeps the roadmap focused.

Turning themes into Jobs to Be Done

Once themes are clear, rewrite them from the user’s point of view.

A helpful framework is Jobs to Be Done (JTBD). It shifts the focus from feature requests to the real goal someone is trying to achieve.

A common format is:

“When [situation], I want to [motivation], so I can [expected outcome].”

Example theme: “People struggle to find past invoices.”

JTBD version:

“When I’m doing monthly accounting, I want to quickly locate and download a specific invoice, so I can close my books without wasting time searching.”

This turns a complaint into a clear outcome. It also helps your team understand what “done” looks like.

If you want more structure for turning findings into roadmaps, this guide can help: product management 101 for media publishers.

Your next steps: from findings to tests

The last step is turning each theme or JTBD statement into a hypothesis or a specific next action.

This is how customer insight connects to the roadmap.

• Finding: “People are confused by our three pricing tiers.”

  Hypothesis: “If we test a simpler two-tier pricing page, trial-to-paid conversion will increase.”

• Finding: “People don’t realize we have a mobile app.”

  Next step: “Add a clear dashboard banner and track clicks to app downloads.”

This step prevents research from becoming “interesting notes” that never change anything. It turns learning into action.

Frequently asked questions

Founders often hit the same practical questions when they start talking to customers. These answers should help you keep momentum.

How many people do I need to talk to?

For interviews and moderated usability tests, patterns often appear quickly.

In many products, you start hearing repeat issues after 5–8 conversations within the same user group.

You are not trying to prove statistics in early discovery. You are trying to find major friction, unmet needs, and the language people use to describe the problem.

What budget should I plan for?

You can do this cheaply, especially at the start.

If you recruit from your list or community and run sessions yourself, your main costs are time and incentives. Many founders start with $25–$50 per participant for a 45-minute session.

If your audience is rare or hard to reach, recruiting costs go up. In those cases, incentives may need to be much higher to get reliable participation.

The cost of skipping research is often bigger. Building the wrong thing is expensive.

What beginner tools should I use?

Start with tools you will actually use every week.

• Scheduling: Calendly
• Video calls and recording: Zoom or Google Meet
• Synthesis: Google Sheets or Miro
• Unmoderated tests: Maze or Lyssna

Tools matter less than consistency. A weekly rhythm of customer conversations beats a perfect stack you never use.

How do I fit this into a fast roadmap?

Do not treat discovery as a big project that stops development.

Instead, make it a small habit. Talk to one or two users each week. Review notes, spot patterns, and feed the best findings into the next sprint.

This keeps learning close to shipping. It also reduces the odds you spend months building in the wrong direction.

---

At Refact, we help founders move from uncertainty to clear product decisions. If you want help planning customer interviews, testing prototypes, and turning findings into a build plan, we can help. Learn more about our approach.