The Universal Prototyping Cycle
Regardless of methodology, every prototyping effort follows the same fundamental loop. The frameworks below differ in emphasis and structure, but they all embody this cycle:
Then repeat — with increased fidelity if the hypothesis holds, or a new hypothesis if it doesn't.
Design Thinking
Developed at Stanford's d.school and popularized by IDEO, Design Thinking is a human-centered framework that emphasizes deep empathy with users before generating solutions. It is widely used in product design, service design, and organizational innovation.
The Five Stages
Empathize
Observe and engage with real users. Conduct interviews, shadow their workflow, and map their pain points. The goal is to understand their needs deeply — not what they say they want, but what they actually struggle with.
Tools: User interviews, contextual inquiry, empathy maps, journey maps.
Define
Synthesize your research into a clear problem statement (Point of View). Frame it from the user's perspective: "[User] needs [need] because [insight]." A well-defined problem is half the solution.
Tools: How Might We questions, POV statements, affinity diagrams.
Ideate
Generate a wide range of possible solutions. Prioritize quantity and creativity over quality. Suspend judgment. Use structured brainstorming techniques to push past obvious ideas.
Tools: Brainstorming, Crazy 8s, SCAMPER, mind maps, dot voting.
Prototype
Build a quick, tangible representation of the best idea(s). Start low-fidelity. The prototype should be cheap, fast, and disposable. Its sole purpose is to make the idea testable.
Tools: Paper prototypes, Figma, cardboard models, role-playing.
Test
Put the prototype in front of real users and observe. Don't explain or sell — watch what they do. Gather feedback, identify what works and what doesn't, and loop back to any earlier stage as needed.
Tools: Usability tests, think-aloud protocol, A/B tests, feedback grids.
Key insight: Design Thinking is not linear. You should expect to loop between stages. Testing often reveals you need to go back to Empathize, or that the problem statement from Define needs refining.
Lean Startup
Popularized by Eric Ries, the Lean Startup method treats every new product as a series of experiments. It is especially effective for startups and new ventures operating under extreme uncertainty, where the business model itself is unknown.
The Build-Measure-Learn Loop
Build
Create the minimum viable product (MVP) — the smallest thing you can build to test your riskiest assumption.
Measure
Collect actionable metrics (not vanity metrics). Focus on data that tells you whether users find value, not just whether they visited.
Learn
Analyze the data and decide: persevere (keep going), pivot (change direction), or kill (stop the experiment).
How-To: Running a Lean Experiment
- Step 1.State your hypothesis: "We believe [target customer] will [expected behavior] because [reason]."
- Step 2.Identify the riskiest assumption: What must be true for this to work? Typically: "Users have this problem" or "Users will pay for this solution."
- Step 3.Define success criteria before building: "If X% of visitors sign up, we proceed." Set the bar in advance to avoid post-hoc rationalization.
- Step 4.Build the smallest possible MVP: Landing page, concierge, Wizard of Oz, or smoke test. Do not build features; build a test.
- Step 5.Run the experiment with real users: Drive traffic, recruit participants, or deploy to a beta cohort.
- Step 6.Measure against your pre-defined criteria: Did you hit the bar? If yes, increase fidelity. If no, analyze why and decide whether to pivot or iterate.
Key concept — Innovation Accounting: Traditional metrics (revenue, profit) are lagging indicators. Use leading indicators — activation rate, retention, task completion — to evaluate whether you are making progress toward product-market fit.
Google Design Sprint
Created at Google Ventures by Jake Knapp, the Design Sprint is a structured five-day process for answering critical business questions through design, prototyping, and testing. It compresses months of work into a single week.
The Five Days
Monday — Map
Define the challenge. Map the user journey end-to-end. Interview domain experts. Choose a target area to focus on for the week. End the day with a clear sprint question.
Tuesday — Sketch
Review existing solutions and inspiration (Lightning Demos). Each team member sketches their own solution independently. This ensures diverse ideas without groupthink. Use the Crazy 8s exercise for rapid ideation.
Wednesday — Decide
Review all sketches in a structured "art museum" walkthrough. Vote on the strongest concepts. The Decider (usually the product owner) makes the final call. Create a storyboard for the prototype.
Thursday — Prototype
Build a realistic-looking prototype in one day. It should look real but doesn't need to work. Divide and conquer: one person makes assets, one writes copy, one assembles in Figma/Keynote. Focus on the critical path from the storyboard.
Friday — Test
Test the prototype with five target users in one-on-one sessions. One interviewer, one note-taker, the rest of the team watches on a live stream. Look for patterns across all five sessions. End the day with clear conclusions.
Why five users? Jakob Nielsen's research shows that five users typically uncover ~85% of usability issues. More users produce diminishing returns for qualitative testing.
Assumption Mapping
Before choosing any framework, it helps to explicitly map your assumptions. This technique works as a standalone method or as a precursor to any of the approaches above.
How-To
- Step 1.List all assumptions that your product/feature depends on. Include desirability ("users want this"), viability ("we can make money from this"), and feasibility ("we can build this").
- Step 2.Plot on a 2x2 matrix: X-axis = evidence (low to high), Y-axis = risk if wrong (low to high). The top-left quadrant (high risk, low evidence) contains your leap-of-faith assumptions.
- Step 3.Design experiments to test the top-left assumptions first. These are the ones that could invalidate the entire idea.
- Step 4.Sequence experiments from cheapest/fastest to most expensive. Don't run an A/B test if an interview could answer the question.
Typical Stages of a Prototyping Initiative
Regardless of which framework you use, most prototyping initiatives go through these phases:
Phase 1: Discovery (1–2 weeks)
Desk research, stakeholder interviews, competitive analysis, user research. Output: clear problem statement and prioritized assumptions.
Phase 2: Ideation & Prioritization (2–5 days)
Generate solution concepts. Score against user impact, feasibility, and strategic fit. Select 1–3 concepts to prototype.
Phase 3: Rapid Prototyping (3–10 days)
Build lightweight prototypes. Start low-fidelity and increase as confidence grows. Multiple small prototypes are better than one large one.
Phase 4: User Testing & Validation (1–2 weeks)
Run structured experiments. Combine qualitative sessions (interviews, usability) with quantitative signals (conversion, engagement). Synthesize findings.
Phase 5: Decision & Next Steps
Based on evidence: proceed to development, iterate with another prototype cycle, pivot the concept, or kill the idea. Document learnings for the organization.
Choosing the Right Approach
| Situation | Best Approach |
|---|---|
| Unclear who the user is or what they need | Design Thinking |
| New venture with unknown business model | Lean Startup |
| Specific feature question, tight timeline | Design Sprint |
| Many unknowns, need to prioritize what to test | Assumption Mapping → then any of the above |
| Optimizing an existing product with data | Lean + Analytics (see next section) |