7 User Testing Steps to Avoid Building a Product Nobody Wants

7 User Testing Steps to Avoid Building a Product Nobody Wants

You’ve got an idea. You’re convinced it’s brilliant. You spend weeks, maybe months, polishing it. You pour your money and soul into building a website or an app. You launch.

And then… nothing.

The reason is almost always the same. You built something you wanted, not what your customers needed. You sat in a room, marvelled at your assumptions, and mistook them for facts. It’s the single most significant, most expensive mistake entrepreneurs make.

The good news? There’s a cure. It's called user testing. And it’s not the complicated, eye-wateringly expensive process some consultants want you to believe it is. It's a simple discipline of watching real people use your creation.

Here’s how you do it without the waste.

Here’s Why You’re Probably Avoiding This

Before we get to the “how,” let's tackle the “why not.” Most people dodge user testing for a few predictable reasons. See if any of these sound familiar.

You're scared they'll call your baby ugly.

This is the big one. Ego. You've slaved over this project, and the thought of a stranger pointing out its flaws feels like a personal attack. You'd rather live in blissful ignorance than face the uncomfortable truth that your brilliant navigation idea is confusing.

Get over it. A bruised ego is cheaper to fix than a bankrupt business.

You think it costs a fortune.

You've heard stories of companies spending tens of thousands on focus groups behind two-way mirrors, with scientists in white coats taking notes. That’s a world away from what you need.

I will show you how to get 90% of the value for less than the price of a decent dinner for two. Sometimes for free. People who want to sell you costly services perpetuate the myth that this is expensive.

The Golden Rule I Beat Into My Clients: You. Are. Not. Your. User.

Repeat it. Burn it into your brain. You know too much. You know what every button is supposed to do. You know the “secret” logic behind the user flow. This knowledge makes you the single worst person to judge your work.

Your opinion is irrelevant. The opinion of your co-founder is irrelevant. The opinion of your partner is irrelevant. The only thing that matters is the confused silence of a real user who can't figure out how to give you their money.

The 7 User Testing Steps That Work

Right, lecture over. Let’s get to the practical stuff. This is a repeatable system. A loop. It's not a one-time exam you pass. It’s a muscle you build.

Step 1: Define the One Thing You Really Need to Know

The first mistake is trying to test everything. You end up with a mess of vague feedback that isn't actionable. You need to focus.

Stop trying to “test the website.” It's too broad. Instead, pick one critical journey or one burning question you have. Frame it as a hypothesis you want to prove or disprove.

• Bad Question: “Do people like our new homepage?”
• Good Hypothesis: “New users can understand our core value proposition and find the pricing page within 30 seconds of landing on the homepage.”

See the difference? One is a useless vanity check. The other is a specific, measurable, and testable objective. You're no longer asking for an opinion but observing a behaviour. Pick one thing. Just one.

Step 2: Find 5 Real Humans (And Not Your Mum)

You don't need 50 users. You don't even need 20.

The godfather of this field, Jakob Nielsen, proved years ago that testing with just five users will typically uncover around 85% of the usability problems in your design. Five. The return on investment after that fifth user drops off a cliff.

So, where do you find these five people?

• Friends of Friends: Your immediate friends know you and will try to be nice. Their friends, however, have less of a filter. This is a great place to start.
• Targeted Social Media Groups: Find LinkedIn or Facebook groups where your target audience hangs out. Post an honest request: “I'm a small business owner looking for 30 minutes of feedback on a new website for [type of person]. I can offer a £20 Amazon voucher for your time.”
• Your Existing Audience (Carefully): If you have an email list or social media following, you can ask for volunteers. The risk is that they are already “superfans” and may be biased.

Sale

Homepage Usability: 50 Websites Deconstructed

• Nielsen, Jakob (Author)
• English (Publication Language)
• 315 Pages – 10/01/2001 (Publication Date) – New Riders Publishing (Publisher)

$50.00 −$23.87 $26.13

Buy on Amazon

Crucially, write a simple screener. These are 2-3 quick questions to get the right people. If you’re testing a new tool for landlords, your screener might ask:

1. Do you currently own and rent out property? (Yes/No)
2. How many properties do you manage? (1-2, 3-10, 11+)

This weeds out the people whose feedback would be useless. And please, don't use your mum. She loves you. She’ll tell you it’s terrific. Her feedback is a lie.

Step 3: Build Something to Test (It Can Be Ugly)

Perfectionism is your enemy here. You are not trying to present a finished masterpiece. You are testing a concept. An ugly thing that works is infinitely more valuable than a beautiful thing that doesn't.

Here’s your hierarchy of options, from good to best:

• Good Enough: Paper Sketches. Seriously. A drawing of a screen on a piece of paper. You can point and ask, “What would you tap here?” It costs nothing; you can test a core idea in ten minutes.
• Better: Interactive Wireframes. Tools like Figma have generous free tiers. You can link static images of your pages together to create a clickable, interactive prototype. It feels real enough to the user, but you haven't wasted a second writing code.
• Use With Caution: Your Live Site. Testing your live, public-facing website is fine, but only if you're testing an existing flow. It’s often too late for new ideas—you’ve already spent the money building it.

Key Insight: One client I worked with spent three weeks designing a “pixel-perfect” prototype for a new checkout flow. It was beautiful. We put it in front of three users. All three got stuck at the same spot and couldn't complete their purchase. The feedback was invaluable, but he could have learned the same from a simple Figma wireframe in a single afternoon. Test the flow, not the font.

Step 4: Write Your Script (But Be Ready to Go Off-Piste)

A script isn't about reading lines like a robot. It's a checklist to ensure you cover the essentials and give every participant the same context. It keeps you on track.

Your script should have four parts:

1. The Welcome: “Hi, thanks for your time. Just to be clear, we're testing the design, not you. You can't do anything wrong. The more you get stuck, the more we learn.” This is the most important thing you'll say. It permits them to be brutally honest.
2. The Context: Briefly set the scene. “This is a new website that helps people find local dog walkers. We're going to look at an early version of it.”
3. The Tasks: This is the core of the session. Don't give instructions. Give scenarios.
   • Bad: “Click on the ‘Services' button, then click on ‘Dog Walking', and then click ‘Book Now'.”
   • Good: “Imagine you're going on holiday next month and need someone to walk your dog, Buster. Using this site, show me how you would find and book a walker for the first week of August.”
4. The Wrap-Up: A few open-ended questions to capture their overall thoughts. “What was the most frustrating part of that? Was there anything that surprised you?”

Straight Talk: Re-read that first point. You are testing the product, not the person. Say it. Let them know that their confusion is your fault, not theirs. This single sentence will unlock the most valuable feedback.

Step 5: Run the Session and Learn to Shut Up

This is where the magic happens, and it’s the most challenging part for most founders. Your job is to facilitate, not participate.

Introduce the Think-Aloud Protocol. It's a simple instruction: “As you go through this, please try to speak your thoughts out loud? Tell me what you're looking at, trying to do, and what you expect to happen. There’s no right or wrong, I’m just interested in your thought process.”

Then, you shut up.

Your instinct will be to jump in and help them when they struggle. “Oh, the button you're looking for is just over there!” Don't do it. Every time you help, you destroy a learning opportunity.

Embrace the uncomfortable silence. When a user pauses, looking confused, don't speak. Count to ten in your head. Let them work through it. The moment they sigh, frown, or mutter “Where the hell is it…?”—that's the gold. That's the friction point you came here to find.

Your only job is to prompt them to keep thinking aloud. If they go quiet, use neutral prompts:

• “What are you thinking now?”
• “What did you expect that button to do?”
• “What's going through your mind?”

Never, ever ask, “Do you like it?” People are programmed to be polite. They'll say, “Yeah, it's nice.” It tells you nothing. Instead, you're observing their behaviour. What they do is the truth. What they say is often just noise.

Step 6: Find the Patterns in the Mess (This Isn't Rocket Science)

After five sessions, you'll have a pile of notes and some recordings. It can feel overwhelming. Don't fall into the trap of creating a massive, colour-coded spreadsheet.

Use the low-tech, high-impact sticky note method.

1. Watch the Recordings: Go through each session recording.
2. One Observation, One Note: Every time you see a key moment—a point of friction, a moment of confusion, a positive surprise, a direct quote—write it on a single sticky note. Be specific. “User couldn't find the contact number.” “User was surprised they had to create an account to see prices.”
3. Stick Them on a Wall: After you've reviewed all the recordings, you'll have a cloud of sticky notes.
4. Group Them: Now, start clustering them. Move the notes around. You'll quickly see patterns emerge. A big cluster of notes about the confusing checkout process. Another little group about the unclear pricing.

This is called affinity mapping. It sounds fancy, but it's just organised common sense. You're not looking for statistical significance. You're looking for recurring themes. If three or four of your five users stumbled in the same place, that's not a coincidence. That's a fire alarm. You have found a problem that needs fixing.

Step 7: Decide, Act, and Do It All Again

Now you have a prioritised list of problems, based on real user pain. The final step is to turn these insights into action.

Don't try to fix everything. Look for the low-hanging fruit and the most critical blockers. That checkout flow that stopped three people from giving you money? That's your number one priority. The minor typo someone noticed in the footer? That can wait.

Make the change. Adjust the design. Rewrite the confusing copy. Then—and this is the part everyone skips—test it again. Maybe not with a full five users, but grab two or three new people and see if you've solved the problem.

This transforms user testing from a “task to be completed” into the engine of your business. It's a loop: Build -> Test -> Learn -> Repeat. It's a mindset of continuous improvement, driven by evidence, not ego.

The Classic, Expensive Mistakes I See Every Single Day

To wrap up, here’s a quick-fire list of the traps that will invalidate all your hard work. Avoid these like the plague.

• Leading the Witness: “The button to book is right here, it's easy, just click it, right?” You've just told them the answer and shamed them into agreeing.
• Testing with the Wrong People: Getting feedback on your high-end financial software from a university student wastes everyone's time. Screen your participants.
• Asking for Opinions: “What do you think of the colour blue we used?” Who cares? Does the design help them achieve their goal or not?
• Getting Defensive: When a user says, “This is confusing,” don't reply with “Well, actually, it's designed to do…”. Just write it down. Thank them. Their confusion is the data.
• Treating It as a One-Off: User testing isn't a final exam you take before launch. It's a regular health check for your product.

A Quick Word on Your Toolkit

You don't need to spend a fortune. Start with what's free.

• The Bare Essentials (Free): A video call tool like Zoom or Google Meet to share screens and record the session. A notepad (digital or physical). That's it.
• The Next Level (Freemium): Create interactive prototypes using a free Figma account. Use a service like Maze to run unmoderated tests where users complete tasks independently.
• The “Pro” Stuff (When You're Scaling): When you have a bigger budget and must constantly test, tools like UserTesting.com or Lookback can help you recruit participants and manage sessions more efficiently. But don't even think about these until you've mastered the basics with the free stuff.

This is a Mindset, Not Just a Task

User testing isn't just another item on your to-do list. It's a fundamental shift in how you build your business. It's the decision to replace “I think” with “I know.” It's the humility to accept that your first idea is rarely your best.

The choice is simple: you can keep building things in the dark, guessing what people want and hoping you get lucky. Or you can switch on the light, watch how real people behave, and build something they will use and pay for.

It's the difference between being an artist and being a designer. An artist creates for themselves. A designer creates for others. If you're in business, you're a designer. Start acting like one.

Getting this raw user data is the first, crucial step. But if your tests reveal a deep, fundamental flaw in your website's architecture or messaging, that's when you need more than a quick fix. That's a foundational design problem.

Our approach to web design services is built on this principle of user-centric clarity. If you're tired of guessing and ready to make something your customers will genuinely thank you for, request a no-nonsense quote.

Look at our other posts on the Inkbot Design blog for more brutally honest advice.

FAQs about User Testing

Last update on 2025-07-21 / Affiliate links / Images from Amazon Product Advertising API