Write a hypothesis
Connect a focused design change to an outcome you can measure.
Treat the design idea like a hypothesis. Be clear about who it is for, what might change for them, and which measure would help you learn whether the change worked.
We believe [a focused change] for [specific people] will lead to [a user outcome], measured by [a metric].
Make it measurable
Avoid goals such as “make the app better.” Choose an observable behavior or outcome, such as fewer errors, less time waiting, or more completed tasks. Set a success measure before testing so you can compare the result with your expectation.
The speaker gives an example of a live map for daily public-transport commuters. The intended outcome is less waiting and more confidence; daily active use and user feedback are suggested measures. It is an example to make the framework concrete, not a target for your own project.
Learn from evidence
The session describes a repeating cycle: form a hypothesis, run an experiment, measure what happened, learn, and repeat. Evidence can change your mind about a design decision. Keep the test small enough that the result teaches you something specific.