The Hallucination of the Builder and the Curse of Knowledge

Engineers, Designers, and Product Managers all suffer from a fatal disease: The Builder’s Bias (also known as the Curse of Knowledge). We know how the feature works because we built it. We know that the “Save” icon is a floppy disk. We know that you have to click “Edit” before you can type in the field. We know that the three dots menu contains the “Delete” option. It is obvious to us. To the user, it is baffling. They look at the screen and see noise. They are afraid to click the wrong button. They don’t know what “Hamburger Menu” means. User Testing is the act of watching a stranger try to use your product without your help. It is often painful. You will watch them struggle to find the massive “Sign Up” button that is right in the center of the screen. You will want to scream: “IT IS RIGHT THERE!”. But you cannot. You must sit on your hands and observe. Because if they can’t find it, it’s not their fault. It’s your fault. User Testing bridges the gap between the “Ideal User” in your head and the “Real User” in the wild.

Why Maison Code Discusses This

At Maison Code, we don’t just write code. We build Products. We have seen startups waste $500k building a complex feature set that nobody understands. They launch. Crickets. “Why aren’t they using the Advanced Filter?” Because they didn’t know it existed. Before we write a single line of code, or immediately after the MVP prototype, we advocate for Usability Testing. We facilitate these sessions effectively. We bring the “Voice of the Customer” into the engineering room. Data is hard to argue with. “The generic opinion is that blue is better.” vs “5 out of 5 users failed to find the checkout button.” The second statement drives action. We talk about this because UX Debt is as real as Tech Debt.

Qualitative vs Quantitative: The Full Picture

You need both data streams to make decisions.

Quantitative Data (The “What”):

Tools: Google Analytics, Mixpanel, Amplitude.
Insights: “Bounce rate is 60% on the Pricing Page.” “Conversion rate dropped by 2% on Mobile.”
Blind Spot: It doesn’t tell you why. Did they bounce because the price is too high? Or because the page didn’t load? Or because the font was ugly?

Qualitative Data (The “Why”):

Tools: User Interviews, Usability Testing, Session Replay.
Insights: “The user didn’t scroll down because the ‘Hero Image’ looked like a full-screen landing page (The False Bottom illusion).” “The user didn’t trust the site because the stock photos looked fake.”
Result: This gives you the specific actionable fix. “Decrease Hero height to 80vh so the next section is visible.”

The Methodology: The Mom Test

We follow the principles of “The Mom Test” by Rob Fitzpatrick. If you ask your Mom: “Do you like my app?”, she will say “Yes, honey, it’s beautiful.” She is lying to protect your feelings. Most users want to be nice. Rule #1: Never ask for opinions (“Do you like it?”). Ask for behaviors. Rule #2: Give them a specific Task (The Scenario). “Imagine you are planning a trip to Paris for next weekend. Try to book a hotel under $200.” Rule #3: Shut up. Don’t guide them. “Click there.” “No, not that one.” If they get stuck, let them struggle for a bit. The struggle is the data. If they ask “What does this button do?”, reflect it back: “What do you think it does?”

Moderated vs Unmoderated Testing

There are two main ways to run this.

1. Moderated Testing (Zoom / In-Person)

You are on the call with the user.

Protocol: “Think Aloud”. Ask the user to narrate their thoughts. “I’m looking for the price… I can’t find it… I’m clicking this menu…”
Pros: Deep insights. You can ask “Why did you hesitate there?”. You can pivot the test if they find a new bug. Empathy building.
Cons: Expensive. Time consuming (1 hour per user + scheduling). Risk of biasing the user (The “Hawthorne Effect” - they act smarter because they are being watched).

2. Unmoderated Testing (UserTesting.com, Maze, Lookback)

You upload a prototype (Figma link) or a live URL. You write a script of tasks. Users record their screen and voice while doing it alone in their bedroom.

Pros: Fast. You can launch a test at 5 PM and have 10 videos by 9 AM. Cheap(er).
Cons: No follow-up possible. Users might phone it in (just clicking fast to get paid).
Strategy: Use Moderated for “Discovery” (What should we build?). Use Unmoderated for “Validation” (Did we build it right?).

Tools of the Trade

Figma Prototypes: The cheapest code is the code you don’t write. Build a “Clickable Dummy” in Figma. It looks real but has no backend. Test this. If users fail, you change the drawing. It takes 10 minutes. If you test the Code version, changing it takes 2 days. fail Fast.
Recruitment Platforms: UserTesting.com: The enterprise standard. Expensive but instant access to specific demographics (“A dentist in Ohio making $100k”). UsabilityHub / Ballpark: Cheaper, shorter tests. Use your own list: Email your newsletter. Offer a $50 Amazon Gift Card. These are your real users.
Passive Testing (Hotjar / Microsoft Clarity): Install a script on your live site. It records video sessions of real visitors (anonymized). Watch 10 sessions while eating lunch. Heatmaps: Show where people click. Rage Clicks: Rapid clicking on an element. Indicates “This looks like a button but isn’t.” -> Fix this immediately.

The 5-User Rule (Nielsen Norman Group)

Jakob Nielsen proved mathematically that testing with 5 users finds 85% of usability problems. You don’t need 50 users.

User 1 finds the critical blockers (Login is broken).
User 2 confirms them.
User 3 confirms them.
User 6 adds very little new info (Diminishing Returns). Strategy: Run small tests often. Test with 5 users. Fix the bugs. Test with 5 users again. Iterative Loop >> Massive Study.

The Skeptic’s View

“We don’t have time. We have a deadline.” Counter-Point: If you launch a confusing product, you will spend months fixing it and handling support tickets. “How do I reset my password?” -> If 100 people ask this, you have a UX bug. A 1-hour user test can save 1 month of engineering rework. User Testing accelerates the viable product, even if it delays the first release by a day. It is a risk reduction strategy.

FAQ

Q: Who should we recruit? A: People who match your persona. If you sell SaaS for Dentists, do not test with Teenagers. Their mental model is different. If you sell Toothpaste (Consumer), test with anyone.

Q: What is A/B Testing vs User Testing? A/B Testing (Split Testing) is Quantitative. “Which button color converts better? Red or Green?” User Testing is Qualitative. “Why did they click Red?” Use User Testing to generate hypotheses. Use A/B Testing to validate them at scale.

Conclusion

Product Management without User Testing is hallucination. Engineering without User Testing is vanity. You are building for people. Talk to them. Watch them. It will humble you. And it will make your product superior. Get out of the building.

Users confused?

If your support inbox is full of “How do I do X?”, or your churn rate is high, your UX is broken. Maison Code conducts professional UX Research and Usability Testing. We recruit the right users, write the scripts, facilitate the sessions, and deliver actionable “Highlight Reels” of the exact moments users get stuck.

Nobody clicking?

We conduct Qualitative User Testing (Moderated & Unmoderated) to uncover the ‘Why’ behind your analytics. Hire our Architects.