UX Design · Lesson 02 / 13
The UX process
The Double Diamond, design thinking and Lean UX explained step by step — what happens in each phase, which methods to use, and how to write a problem statement the whole team can rally around.
5 min read · 5 quiz questions
Why a process matters
Without a process, teams tend to jump straight from an idea to a solution: someone says "let's add a dashboard", and a week later there is a dashboard. Whether anyone needed it is discovered only after launch.
A UX process slows down the beginning on purpose. It makes sure you understand the problem before committing to a solution, and it builds in moments to test your assumptions while changes are still cheap.
There is no single official UX process. Most models share the same idea: understand → define → explore → test → deliver, repeated as many times as needed. This lesson covers the three you will hear about most.
The Double Diamond
The Double Diamond was published by the UK Design Council in 2005 after studying how design teams at large companies worked. It has two diamonds. In each, you first diverge — open up and explore widely — and then converge — narrow down and decide.
1. Discover (diverge)
Explore the problem space with an open mind. You are not designing yet — you are learning.
- Stakeholder interviews to understand goals and constraints.
- User interviews, observation and contextual inquiry.
- Analytics review, support ticket analysis, competitor review.
2. Define (converge)
Make sense of everything you learned and decide which problem to solve.
- Affinity mapping to find patterns in research.
- Personas, journey maps and key insights.
- A clear problem statement and success metrics.
3. Develop (diverge)
Generate many possible solutions to the defined problem. Quantity first, quality later.
- Sketching, crazy 8s, brainstorming and co-design workshops.
- User flows and rough wireframes for several directions.
- Quick prototypes to compare ideas.
4. Deliver (converge)
Test, refine and ship the strongest solution.
- Usability testing and iteration.
- High-fidelity design and developer handoff.
- Launch, measure and learn.
Live preview
The point between the two diamonds — the problem statement — is the most important artefact of the whole process. If it is wrong, everything after it solves the wrong problem.
Design thinking
Design thinking, popularised by IDEO and the Stanford d.school, describes the same journey in five modes. They are called modes, not steps, because you move between them freely.
| Mode | Goal | Typical activities |
|---|---|---|
| Empathize | Understand people's needs, feelings and context | Interviews, observation, immersion |
| Define | Frame the right problem | Synthesis, point-of-view statements, How Might We questions |
| Ideate | Generate many ideas | Brainstorming, sketching, mind maps, SCAMPER |
| Prototype | Make ideas tangible | Paper prototypes, role-play, clickable mockups |
| Test | Learn what works | Usability tests, feedback sessions |
Lean UX and working in sprints
In many product teams, design happens alongside agile development in one- or two-week sprints. Lean UX, described by Jeff Gothelf and Josh Seiden, adapts the process to that pace. Instead of long research phases and big deliverables, it focuses on assumptions, hypotheses and fast experiments.
- 1List your assumptions about users, problems and solutions.
- 2Turn the riskiest ones into testable hypotheses.
- 3Design the smallest thing that can test the hypothesis (a minimum viable product or experiment).
- 4Run the experiment and measure the result.
- 5Learn, then keep, change or drop the idea.
We believe that [showing delivery fees on the restaurant list]
for [price-sensitive customers]
will achieve [fewer abandoned checkouts].
We'll know we're right when we see [checkout abandonment drop by 10%].Writing a strong problem statement
A problem statement describes the user, their need and the insight behind it — without prescribing a solution. A common format from design thinking is the point-of-view (POV) statement, followed by How Might We (HMW) questions that turn the problem into an invitation for ideas.
Do
- Center the statement on a specific user and a real need from research.
- Include the insight — the "because" that explains the behaviour.
- Keep HMW questions broad enough for many ideas, narrow enough to act on.
Don't
- Put the solution in the problem: "Users need a dashboard".
- Write about the business only: "We need more sign-ups".
- Make it so broad it's meaningless: "How might we improve finance?"
POV:
New freelancers [user] need a way to send their first invoice
quickly [need] because they don't know which tax details are
legally required and fear looking unprofessional [insight].
How might we…
- help new freelancers send a first invoice in under five minutes?
- make tax details feel safe instead of scary?
- let freelancers look professional from day one?Choosing how much process you need
The process should match the risk. Redesigning the payment flow of a banking app deserves serious research and testing. Changing the wording of a tooltip does not.
| Situation | Suggested approach |
|---|---|
| New product or major feature, high uncertainty | Full Double Diamond with real discovery research |
| Improving an existing flow with known problems | Short discovery (analytics + 5 interviews), then design and test |
| Small UI change, low risk | Design, quick review, ship and monitor |
| Fast-moving startup with many unknowns | Lean UX: hypotheses and experiments every sprint |
Common mistake: Treating the process as a checklist of deliverables. Personas and journey maps are only valuable if they change decisions. If nobody uses an artefact, it was the wrong one.
Key takeaways
- UX processes all follow the same idea: understand the problem, define it, explore solutions, test and deliver.
- The Double Diamond alternates diverging (exploring) and converging (deciding) — first for the problem, then for the solution.
- Design thinking's five modes — Empathize, Define, Ideate, Prototype, Test — are a loop, not a straight line.
- Lean UX turns assumptions into hypotheses and tests them quickly inside agile sprints.
- A good problem statement names the user, the need and the insight — and contains no solution.
Exercise
Write a problem statement
- Choose a frustration you or a friend had recently with a digital product.
- Write a POV statement: [user] needs [need] because [insight].
- Check it: does it contain a solution? If so, rewrite it.
- Write five How Might We questions from your POV.
- Circle the one that would lead to the most interesting ideas.
Test yourself
The UX process quiz
5 questions. Score 70% or more to complete the lesson.
