Prototyping & Usability Testing
Test the flow while changing it is still cheap.
Interactive prototypes, realistic scenarios and observed tasks that produce findings you can iterate on before engineering spends a quarter implementing the wrong path.
Validation before build
A prototype is a question. A test is the answer.
We build the smallest interactive prototype that can hold the risky flow, write scenarios and tasks, then watch people attempt the work. Findings go back into the design while the cost of change is still a file, not a release. The point is risk reduction, not a theatre of high-fidelity screens nobody was asked to use.
Typical constraints
The build is about to freeze a flow nobody has tried.
- Stakeholders approved static frames and called that validation.
- The risky journey has never been tapped or clicked through by a stranger.
- Engineering estimates assume the current path is correct.
- Previous tests produced quotes and no design changes.
- Fidelity is so high that teams are arguing about shadows instead of the task.
- A mobile or multi-step flow cannot be judged from a slide deck.
How testing is run
Question, prototype, task, finding, change.
We match fidelity to the risk. A navigation question does not need production visual design. A gesture question might.
01
Choose fidelity for the question
What must be interactive, and what would only distract the participant.
02
Script scenarios and tasks
Jobs a real person would recognise, with success criteria the team agrees on.
03
Build the interactive prototype
The risky path and the branches that usually hide the failure.
04
Observe people attempting the work
Facilitation that watches behaviour, not a focus group about taste.
05
Capture findings without theatre
Severity, frequency and design implication, without inventing a statistic.
06
Iterate before build lock-in
The flow changes while it is still cheap, then we retest if the risk remains high.
What you receive
A prototype, a test, and a cheaper next version.
The engagement is a loop: prototype, observe, change. You leave with findings and an iterated flow, not only a recording archive.
01
Question and fidelity choice
What we need to learn, and the lowest fidelity that can answer it.
02
Interactive prototype
The critical path in a form people can actually attempt, including key branches.
03
Scenarios and tasks
Realistic jobs, not 'please click around and tell us what you think'.
04
Observed sessions
Facilitated tests with notes on where people hesitate, recover or fail.
05
Findings and severity
What blocked the task, what was merely disliked, and what can wait.
06
Iterated flow
The design updated from the evidence before the build path is locked.
Useful when
This is the brief when being wrong in production would hurt.
- A mobile app is about to enter native build on an untested core loop.
- A checkout, onboarding or permissions flow has too many branches to judge on paper.
- A design system change might break tasks people already know.
- Stakeholders disagree and a watched task would settle the argument.
- Research already explained the need, and the remaining risk is in the interaction.
- Engineering wants confidence before a sprint commits to the current path.
What we judge
The expensive mistakes should happen in the prototype.
A flow people can complete
The task was attempted by someone who did not design it.
Findings that changed the file
The prototype is not a prop. The next version exists because of the sessions.
Lower build risk
Engineering starts from a path that has already failed once, cheaply.
A clearer argument in the room
Taste debates lose to watched behaviour when the two disagree.
Related work
Selected prototype and test work will appear here.
Usability sessions are usually confidential. We will not invent task-success rates or quotes. Approved studies will be added to the Work index when they can be shown.
Published case studies will appear here when they are cleared.
Related services
Work that usually sits beside this.
Questions
Buying questions, answered directly.
Related, not the same. Research is broader: interviews, mapping, synthesis. This work is specifically about interactive prototypes and observed tasks on a flow you are about to build.
Next move
If the flow is expensive to get wrong, prototype it.
Tell us the journey you are about to build and what would hurt if it failed. We will say whether a prototype test, broader research or a design-system pass is the useful move.