Usability Testing

Why is Usability Testing Important?

Imagine spending an entire month designing an app, and after it was developed and launched you discovered that users couldn't find the purchase button. That problem could have been uncovered in 30 minutes with a simple usability test.

Usability Testing is watching real users as they try to complete specific tasks in your product. You're not asking for their opinion (that's something different) — you're observing them actually using the product.

Steve Krug — author of Don't Make Me Think — says that testing with even a single user is a million times better than no testing at all. And the truth is that most major problems are revealed by the first 3 to 5 users. You don't need a large budget or a research team — you just need to start.

Jakob Nielsen conducted a famous study showing that 5 users are sufficient to uncover 85% of usability problems. So you don't need to test with 50 or 100 people — 5 is enough if you choose them correctly and ask the right questions.

Planning the Test

Before you start testing, you need to plan well. Poor planning produces unusable results.

1. Define test objectives
What do you want to learn? Examples:

• Can users complete the registration process in under 3 minutes?
• Can they find the search button easily?
• What is the most common step where users get stuck in the checkout process?

2. Define participants
Who will test? They must be representative of real users. If you're designing an app for mothers, test with actual mothers. If you're designing a tool for developers, test with developers.

Participant selection criteria:

• Age and gender
• Technical level
• Experience with similar products
• Any important characteristics for your target audience

3. Prepare the tasks
Tasks are the things you'll ask the participant to do. Good tasks are:

• Realistic and specific: "Buy a blue T-shirt in size L" is better than "buy something"
• Without hints: "Find a way to change the delivery address" is better than "click on Settings then Address"
• Related to test objectives
• Ordered from easy to hard

4. Prepare the environment
Whether testing online or in person, you need to prepare:

• The prototype or product to be tested
• Screen and audio recording tool
• Consent form
• Notes or follow-up questions
• A way to compensate participants (e.g., a gift card)

Moderated Testing

In moderated testing, you are present with the participant — either in person or via video call — and guide them through the test.

Advantages of moderated testing:

• You can ask follow-up questions in real time
• You can clarify the task if the participant misunderstood it
• You can see facial expressions and body language
• You can go deeper into specific points

Session management steps:

Introduction phase (5 minutes):

• Welcome the participant and introduce yourself
• Explain that you're testing the product, not them — there's no right or wrong answer
• Ask them to think aloud (Think Aloud Protocol) — say everything that comes to mind as they use the product
• Get their consent for recording

Task phase (20-30 minutes):

• Present tasks one at a time
• Let the participant try on their own — don't help unless they're completely stuck
• Take notes: where do they hesitate? Where do they make mistakes? What are they saying?
• Ask follow-up questions: "What were you looking for?" "What did you expect to happen?"

Discussion phase (10 minutes):

• Ask about the overall impression
• Return to specific points you noticed and ask about them
• Ask if there's anything they'd like to add

Important tips during facilitation:

• Don't say "that's easy" or "you're supposed to do this" — let the participant discover on their own
• Don't defend the design if the participant criticizes it
• Keep your face neutral — don't show that you're pleased or frustrated
• If the participant asks "should I press here?" say "What do you think? What feels logical?"

Unmoderated Testing

In unmoderated testing, the participant completes the test on their own without anyone present.

Advantages of unmoderated testing:

• Much faster — you can test with a large number in a short time
• Cheaper — you don't need to reserve your time for each participant
• Results closer to reality — the participant uses the product in their natural environment
• Can test with participants from different countries

Disadvantages of unmoderated testing:

• You can't ask follow-up questions in real time
• Some participants may not complete the test
• Response quality may be lower
• You won't see facial expressions

When to use each type?

• Moderated testing is better in early stages when you're trying to understand "why" the user does a certain thing
• Unmoderated testing is better when you want to test a specific thing with a large number and get numbers (how many people completed the task? how long did it take?)

Writing Effective Test Tasks

Tasks are the heart of the test. A poorly written task produces unusable results.

Task writing principles:

1. Make them realistic
Instead of: "Use the filtering feature"
Write: "You want to buy a winter jacket for under $50. Find the right jacket"

2. Don't hint at the answer
Instead of: "Click on the settings icon and change the language"
Write: "You want to change the app language to English. How would you do that?"

3. Define the success criterion
For each task, define in advance: what counts as "success"? Does the participant need to complete the task in a certain time? Without errors? Or just reaching the end?

4. Start with easy tasks
Begin with a simple task so the participant feels comfortable, then gradually increase difficulty.

Examples of good tasks:

• "A friend told you about this app and you want to register. Start from here"
• "You bought a product and want to return it. What are the steps?"
• "You want to know how much time is left on your delivery"
• "Someone sent you a message and you want to reply"

Analyzing Test Results

After finishing the tests, you have a large amount of data. How do you analyze it?

1. Start with immediate observations
Immediately after each session, write down the top 3-5 observations. Immediate observations are more accurate than what you'll remember a week later.

2. Create a problem table
Create a table with:

• Problem: clear description
• Severity: high (couldn't continue) / medium (took a long time) / low (noticed but not affected)
• Frequency: how many participants encountered the problem
• Location: where in the product did the problem occur

3. Look for patterns
A problem encountered by one participant may be a coincidence. A problem encountered by 3 of 5 participants is a real problem that needs to be fixed.

4. Prioritize
You won't be able to fix everything at once. Prioritize problems by:

• Severity × Frequency = Priority
• High severity and high frequency problems are fixed first

5. Suggest solutions
For each problem, suggest at least one solution. The solution doesn't have to be perfect — what matters is that it's a step forward.

Usability Testing Tools

Unmoderated testing tools:

Maze: One of the most well-known tools. Integrates directly with Figma. You can define tasks, see Heatmaps for each screen, know where people tap, and see success rates.

UserTesting: A comprehensive platform that provides participants from a large database. Records videos of participants using the product. A bit expensive but very powerful.

Hotjar: Provides Heatmaps and Session Recordings for real websites. Meaning you're not testing a prototype, you're watching real users on the actual website.

Lookback: Specialized in remote moderated testing. Provides a video call with screen sharing and recording.

Free or affordable tools:

Google Forms + Loom: You can send a prototype link along with a Google Form containing the tasks, and ask the participant to record their screen with Loom.

Figma Prototyping: Figma itself has a feature where you can create a prototype, share a link, and see basic statistics.

Useberry: Cheaper than Maze and offers similar features. Suitable for small teams.

Other Types of Usability Testing

1. A/B Testing
You show two versions of the same design to two different groups and see which one achieves better results. For example, a green button vs. a blue button — which one do people tap more?

2. Card Sorting
You ask participants to organize items into logical groups as they see them. Very useful for designing Information Architecture and Navigation.

3. Tree Testing
You test the Navigation structure without a visual interface. You give the participant a task and they navigate through the section tree to find the right place. Reveals whether the structure is logical or not.

4. First Click Testing
You show a screen and ask the participant: "Where would you click to do X?" Studies show that if the first click was correct, the likelihood of the user completing the task successfully increases by 87%.

5. Five-Second Test
You show a screen for only 5 seconds then ask the participant: "What do you remember? What's the purpose of this page? What was the first thing that caught your eye?" Useful for testing visual hierarchy and main messages.

Writing the Test Report

After analyzing the results, you need to communicate them to the team in a clear and convincing way.

Report components:

• Executive summary: key results on one page — this is what managers will read
• Methodology: how many participants, what tasks, what tools
• Detailed findings: each problem in detail with screenshots and videos
• Recommendations: suggested solutions ranked by priority
• Next steps: what needs to happen next

Tips for the report:

• Use video clips from the sessions — these are more powerful than any words
• Make results specific and actionable
• Rank problems clearly (high/medium/low)
• Present the report in a meeting rather than just sending an email

Conclusion

Usability testing is not a luxury and doesn't require a large budget. Even if you run a simple test with 3 friends, you'll discover problems you didn't know existed.

The golden rule: test early and test often. Don't wait until the design is perfect — test the wireframes, test the prototype, and test the final product. Each stage will give you different and valuable information.

Start simple: take your prototype and have someone use it in front of you. Watch where they get stuck, and listen to what they say. This single step will transform how you design for the better.