Most teams have one idea and call it a decision. The first solution that occurs to you is not the best one — it is the most available one, borrowed from the last product you happened to use.
Ideation buys you options. The rule is that generating and judging are separate activities, done at separate times, with a visible line between them. Judgement contaminates generation: the moment someone says "engineering will never build that", the room starts self-censoring and you are left with the safe list.
Go wide on purpose, then narrow deliberately against the problem statement and the real constraints. A team that skips the wide part has not chosen a solution. It has accepted the first one it met.
- A Systematic Approach to New Product Ideation — article
- Stage 3 in the Design Thinking Process: Ideate — article
Brainstorming
Unstructured group brainstorming is worse than it looks. The loudest person anchors the room and you leave with one idea and six people who believe they chose it.
Structure fixes it. Circulate the problem beforehand, then have everyone write silently for ten minutes, pass their sheet on and build on what they were handed. That is brainwriting, and it beats talking first because nobody gets talked over on paper.
Set a quota, per person: ten ideas each, not "some ideas". A number drags you past the obvious handful to where the interesting ones live.
Ban evaluation while you generate. Not "that is expensive", not "legal will refuse" — write it down and keep going. Filtering is the next session, against criteria you agree before seeing the list: does this solve the problem as written, what would it cost to find out. Criteria chosen afterwards are preferences with arithmetic on top.
- The 6 Key Components of an Effective Product Brainstorming Session — article
- How I Run Brainstorming Sessions (UX Framework) — article
Mind Maps
A mind map puts the problem in the centre and branches outward — causes, user types, moments in time, existing workarounds — and then branches again under each of those.
Reach for it when a linear list is flattening something that is not flat. Forty ideas in a column hide the fact that thirty-six of them are variations on three approaches and four are genuinely different. The map makes that shape visible in a second.
The useful move is inspecting the thin branches. A branch with one lonely child is usually an area nobody has thought about, not an area with nothing in it. Send people back there deliberately.
Build it on a wall or a shared canvas. A mind map is for looking at together, not for filing.
- Gray Matter: What Is a Mind Map in the Design Process? — article
- ES Skip navigation mind maps product design Create Avatar image Design Any Product with this Mind Map — video
Sketching
Pen, paper, a minute a screen. The point of a sketch is not the drawing — it is that the drawing is cheap enough to throw away without regret.
The technique worth learning is Crazy Eights: fold a sheet into eight boxes and fill every one in eight minutes. As first designed, the eight are variations on one promising idea; while you are still going wide, use the same clock for eight different ideas. By box five the obvious moves are spent and you have to invent something.
Sketch alone, then show. A group sketching together produces one shared idea; six people sketching separately produce six, and the differences are the material worth discussing.
Keep it ugly. A neat sketch attracts feedback about the sketch. A rough one gets feedback about the idea, which is what you asked for.
Wireframing
A wireframe answers one question: what is on this screen, and what matters most on it? Boxes, real labels, no colour, no imagery, no font choices.
Low fidelity is a deliberate constraint, not a lack of skill. Show a polished mockup and the conversation is about the shade of blue. Show a grey wireframe and it is about whether the pricing table belongs above the fold.
Use real words. Lorem ipsum hides that your button label is eleven characters too long and that nobody has written the error message yet. Wireframes with fake content pass reviews and then fail in build.
Draw the empty state, the error state and the too-much-data state. The happy path is the part you have already solved; the design breaks on the empty list, the name that will not fit, the payment that fails at the last step.
- What is wireframing? — article
- Figma Wireframe Tutorial for Beginners — video
Prototyping
A prototype is the design made clickable so you can watch someone fail at it before an engineer builds it.
Match fidelity to the question. Testing whether the flow makes sense? A clickable chain of wireframes is plenty. Testing whether people trust your checkout? It has to look real, because trust is partly visual.
Prototype only the narrow path you intend to test and leave everything else as dead links. Wiring every branch is how a two-day prototype becomes a three-week one that nobody dares change, and a prototype too precious to change has become a spec.
Hand it over with a task, not a tour, and then say nothing. The moment you start explaining the prototype you have stopped testing it and started demonstrating it, and a demo cannot fail.
- What is prototyping — article
- UX Prototyping Tutorial - What is a prototype? — video