Choose Your Testing Method
You now know the different ways to test, but which one should you start with? The answer depends on what you’re building and what you need to learn first.
If you’re testing basic market interest and don’t have anything built yet, start with a landing page. This answers the question “do people want this at all”. It’s fast, cheap, and gives you a clear signal. If nobody signs up for your waitlist, you have a problem to solve before building anything.
If you already know there’s interest but need to test the user experience or physical design, use prototypes. For digital products, create a clickable prototype in Figma. For hardware, build a cardboard mockup or 3D model. This shows if people understand how to use it and if it feels right.
If you need to test actual usage and behavior, build a no-code MVP for digital products or a 3D printed prototype for hardware. Or use concierge MVP where you deliver the service manually. This tells you if people will actually use it in real life and if they’ll pay for it.
Create Your Test Plan
A good test plan answers five questions: what are you testing, who will you test with, how long will it take, how much will it cost, and what does success look like.
Start with what you’re testing. Be specific. Not “my app idea” but “whether parents will sign up to find babysitters through a landing page”. Not “my product” but “whether the size and shape of my smart home device feels right to users”. The more specific you are, the better your test will be.
Decide who you’ll test with. Write down exactly who your ideal test participant is. Not just “parents” but “working parents in urban areas with children under 5 who currently struggle to find childcare”. Be this specific. Then figure out where to find them. Which Facebook groups do they hang out in? Which LinkedIn communities? Which local events? Make a list of at least 5 places you’ll recruit from.
Set a realistic timeline. For landing pages, plan 1-2 weeks total. A few days to create the page, a week to run ads and collect data, then time to analyze. For prototypes whether digital or physical, plan 2-3 weeks. One week to build, one week to test with users, time to analyze feedback. For working prototypes or extensive beta testing, plan 4-8 weeks because building and shipping takes time.
Budget realistically. For landing pages, €50-100 covers the page tool and some ad spend. For digital prototypes, €0-50 since tools like Figma are free. For hardware mockups, €50-200 for materials. For 3D printing, €200-800 including iterations. For working hardware prototypes, €1000-3000 depending on complexity. Always add 30% buffer for unexpected costs because something always costs more than expected.
Define success before you start. Write down numbers. For landing pages, success might be 5% conversion rate. For prototype tests, success might be 7 out of 10 users completing the main task without help. For beta tests, success might be 50% of users returning daily for a week. For pre-orders, success might be 15% of people paying a deposit. Having these numbers written down means you’ll know definitively whether your test worked or not.
Prepare Your Materials
Good preparation makes testing run smoothly and gives you better data.
For landing page tests, write your headline focusing on the main benefit not features. Draft your problem statement in your customers’ own words from interviews or research. Create your solution description in 2-3 clear sentences. Design your call-to-action button with action words. Set up email collection asking only for email and optionally name. Choose real photos of real people if possible, not stock photos. Set up Google Analytics before launching so you can track where visitors come from and what they do.
For prototype tests whether digital or physical, use realistic content not placeholder text. If it’s a babysitter booking app, create profiles for realistic babysitters with real names and descriptions. If it’s a physical product, make the mockup actual size with real materials. Write your test script starting with how you’ll introduce the purpose. List the 3-5 main tasks you want them to complete. Prepare your follow-up questions. If testing remotely, test your screen sharing and audio beforehand.
For hardware tests, build or ship prototypes early so testers have time to actually use them. Two weeks minimum for meaningful data. Include clear setup instructions but also watch if they read them or try to figure it out themselves. Both behaviors tell you something important. Create a simple feedback form they can fill out daily or weekly with specific questions about usage.
Track What Matters
You need to track both numbers and observations. Numbers tell you how many people are interested. Observations tell you why they are or aren’t interested.
For landing pages, track visitors, email signups, conversion rate, where traffic came from, and time spent on page. Check these daily during your test. If conversion is very low after 50 visitors, iterate sooner. If it’s doing okay, wait for 100-200 visitors before making changes.
For digital prototypes and MVPs, track daily active users, how long sessions last, which features get used most, task completion rate, and where users get stuck or give up. Most no-code tools have built-in analytics. Set up tracking for key actions like sign up, complete booking, make payment.
For hardware, track usage frequency through surveys or follow-up interviews, how long each session lasts, what breaks or fails, and how many people would actually pay. You’ll need to ask testers to self-report much of this data.
Beyond numbers, take detailed notes during every test session. Capture exact quotes when testers say something insightful. Note when they get confused even if they don’t say anything. Record their facial expressions during key moments. After each session, organize notes by theme like “confusing features” or “missing functionality” or “positive reactions”. This makes patterns easier to spot across multiple testers.
Execute Your Test
Follow your plan but stay flexible enough to learn as you go.
Launch your test and start collecting data. For landing pages, begin with small ad budgets like €10 per day and test different traffic sources. Monitor results daily but don’t change things too quickly. For prototype tests, schedule sessions with clear start and end times and stick to them. Begin each session the same way using your script so you can compare results.
The hardest part is staying quiet during tests. Let testers explore and struggle. Don’t interrupt or help unless they’re completely stuck for over a minute. Your job is to observe not to make them successful. The struggles are where you learn the most. After they complete tasks, then you can ask follow-up questions and dig deeper into their thinking.
Watch for testers who stop using your product or prototype. This is important data. Reach out and ask why they stopped. Was it broken? Not useful? Too complicated? These insights are often more valuable than feedback from people who loved it.
Common Mistakes
Learn from what trips up other founders.
Don’t test with friends and family unless they’re genuinely your target customer. Your mom’s encouragement tells you nothing useful. Don’t test with people who are too polite to give critical feedback. You need honest reactions not encouragement.
Don’t ask leading questions like “wouldn’t it be great if you could do X in 2 minutes”. This tells them what answer you want. Instead ask open questions like “how do you do X today and what’s frustrating about that”. Let them describe problems in their own words.
Don’t dismiss criticism as “they just don’t get it”. If multiple people don’t get it, your product isn’t clear enough. Don’t assume you can teach users how to use it later. Products should be intuitive. If it requires teaching, the design needs work.
Don’t change everything after one test session. Look for patterns across multiple sessions. One person’s opinion might be an outlier. Three people saying the same thing is a pattern. Don’t add features just because someone requested them. Ask why they want that feature and what problem it would solve. Often there’s a simpler solution.
Don’t stop testing after one method. A landing page with good signups doesn’t guarantee people will use the actual product. Prototype tests don’t guarantee people will pay. Each method answers different questions. Use multiple methods to reduce risk before committing to full development.
Summary
Choose your testing method based on what you need to learn most urgently. Create a specific plan with clear success criteria written down as numbers. Prepare all materials before you start so tests run smoothly. Track both quantitative data like conversion rates and qualitative observations like user confusion. Execute your tests systematically while staying quiet and letting users struggle. Avoid common mistakes like testing with friends or asking leading questions. And use multiple testing methods because each one teaches you something different about whether people want and will pay for what you’re building.