Usability Testing

← Back to Method Diagram

“The user’s mental model is not the same as the designer’s”
– Don Norman : NN/g and godfather of UX

Purpose of Usability Testing

Usability testing is a user-centred evaluative research method that observes people completing tasks to identify usability issues and improvement opportunities in a product.

Research tells us what to fix. Usability testing validates we fixed the right thing in the right way – ideally before expensive and otherwise disastrous coding of the ‘solution’ begins.

You sometimes hear Usability Testing referred to as “User Testing”, but this is a misnomer – we are not testing the user, we are testing the usability with the user. 

Usability Testing is still research, even though it comes in the prototyping hypothesis stage of the project.

When we have all the research we need to hypothesise and prototype a solution, it is critical that we usability test our manifest hypothesise (the prototype usually) to ensure that it is right, that the user feels we have solved their problems and that these solutions are usable, intuitive and accessible.

NB. Business’s opinions of what’s good, bad or wrong are a dangerous substitute for real user testing.

“When people use something, they should be able to figure it out without instructions.”
– Don Norman : NN/g and godfather of UX

---

Value of Usability Testing

Usability testing reveals design and usability issues that may not be visible internally and provides direct evidence from target users, supporting more effective UX decisions.

• Validate ideas: Test prototypes early to confirm design decisions before committing to build and release – reduces wasted build time and misdirected development cost.
• Identify usability problems: Direct user task observation exposes navigation, labelling, and interaction failures teams often miss – improves task success rates and reduces support, training, and rework cost.
• Understand user behaviour: Watching real users complete tasks shows how the product is actually used, not how it was intended to be used – increases solution effectiveness and decision accuracy.
• Reduce cost and wasted development effort: Finding and fixing issues before launch avoids expensive late-stage rework – shortens delivery cycles and lowers remediation cost.
• Increase delivery confidence: Evidence from user testing supports design decisions – reduces stakeholder risk and strengthens go-live decisions.
• Strengthen competitive position: Products that are easier to use complete tasks faster and with fewer errors – improves adoption, completion rates, and comparative performance.

---

Anatomy of Usability Test

How to conduct usability testing : A step-by-step framework

There are six core steps to run a usability test:

1. Define the goals of your study

Set clear objectives and success measures. For example, for an e-commerce checkout you might measure task success rate, time on task, and error frequency to evaluate how effectively users can complete a purchase.

2. Write tasks and a script

Create realistic, bias-free tasks that reflect key user goals without revealing the solution path. Focus on what users should be able to accomplish, and prioritise critical functions.

Example: “You need a dress for a wedding. Choose one, select your options, and place the order.”

3. Recruit participants

Recruit representative users via direct outreach, panels, or specialist services. For formative qualitative testing, start with ~5 participants per key user group and increase if the product, journeys, or audiences are complex.

4. Conduct your usability test

Run moderated (facilitated) or unmoderated sessions. Provide scenarios and tasks, and capture task outcomes, errors, decision paths, and observed friction points.

5. Analyse results

Evaluate task success, errors, time on task, breakdown points, and recurring patterns. Frame findings around what the solution enabled or prevented users from doing, and where the design supported or blocked task completion. Prioritise issues by frequency and severity.

6. Report your findings

Summarise findings against the original goals. Present evidence, issue severity, and recommended changes, with clear next steps for design and delivery.

---

Examples

Imagine you’re testing an e-commerce app’s checkout process. You observe several users as they attempt to purchase something, and in the process, uncover issues like unclear form fields and confusing payment options. Based on these observations, you can improve the checkout process by simplifying the language used in the forms and presenting the payment options more clearly.

NB. For security clearance and NDA reasons, some elements of the example files may be redacted, changed or removed.