By the end of this lesson you can
- Write design hypotheses that can be proven wrong.
- Prioritise ideas using impact and effort.
- Choose which assumption to test first and how cheaply to test it.
The idea
Once you have a problem and some ideas, you need to decide what to try first. Treating ideas as hypotheses rather than certainties helps teams learn faster and waste less.
Design hypotheses
A useful format: We believe [change] for [users] will result in [outcome]. We will know we are right when [measurable signal].
"We believe showing delivery slots before checkout for first-time customers will reduce basket abandonment. We will know we are right when abandonment at the delivery step falls by 15% over four weeks." A hypothesis must be possible to disprove; if no result could make you change your mind, it is a plan, not a hypothesis.
Prioritising
A simple impact / effort matrix sorts ideas into four groups: quick wins (high impact, low effort), big bets (high impact, high effort), fill-ins (low impact, low effort) and things to drop (low impact, high effort). Other frameworks, such as RICE (reach, impact, confidence, effort), add more nuance. All of them are tools for conversation, not formulas that decide for you.
Test the riskiest part cheaply
For each big idea, ask what must be true for it to work, and test the riskiest part first with the cheapest method: a paper prototype, a fake door, a conversation or a quick prototype test.
Why it matters in real projects
Teams always have more ideas than time. Hypotheses make the expected outcome explicit, so you can tell whether a change worked. Prioritisation keeps effort on the work most likely to matter. Testing risky assumptions early stops months of building something nobody wanted.
A realistic example
A recipe app team has twelve ideas to increase the number of recipes people actually cook. They score each for impact and effort in a 30-minute session with design, product and engineering.
Two quick wins stand out: a shopping list generated from a recipe, and an "ingredients you already have" filter. The big bet is a step-by-step cooking mode with timers. They write a hypothesis for the cooking mode and identify the riskiest assumption: that people will use their phone while cooking with messy hands.
Before building anything, they film five people cooking at home with a clickable prototype. Most prefer voice and large buttons, which changes the design significantly. The test cost two days; building the wrong version would have cost two months.
Step by step
List ideas
Collect ideas from your HMW sessions in one place.
Write hypotheses
For the main ideas, use "We believe... will result in... We will know when...".
Score impact and effort
Do this as a group with engineering, so effort estimates are realistic.
Place on the matrix
Sort into quick wins, big bets, fill-ins and drops.
Find the riskiest assumption
For each big bet, ask what must be true for it to work.
Plan the cheapest test
Choose the fastest way to learn whether that assumption holds.
Common mistakes
Mistake Hypotheses with no measurable signal.
Instead Say what you will observe or measure and what result would prove you wrong.
Mistake Designers estimating effort alone.
Instead Score effort with engineers, who know the real cost.
Mistake Treating framework scores as final answers.
Instead Use scores to start the discussion, then apply judgement.
Check your understanding
Choose an answer, then check it. There is an explanation either way.
Practice exercise
Exercise · about 45 min
Prioritise and plan a test
Take at least eight ideas from an HMW session (yours or invented for a parking app) and prioritise them.
Steps
- Write a hypothesis for your top three ideas.
- Place all ideas on an impact / effort matrix.
- Identify the riskiest assumption in your biggest bet.
- Describe the cheapest test that would check it.
Deliverable
Matrix, three hypotheses and a short test plan.
Self-assessment
Tick each item when it is true of your exercise. Be honest; this is for you.
4 items
Summary
- Treat design ideas as hypotheses with measurable signals.
- Impact / effort matrices help prioritise and start conversations.
- Test the riskiest assumption first, as cheaply as possible.
- Frameworks support judgement; they do not replace it.
Further reading
Third-party sources, chosen for reliability. They open in a new tab.
Your notes
Marking complete updates your course and roadmap progress.