Background
In November 2022, the Japanese government announced its "Startup Development Five-Year Plan," allocating roughly one trillion yen in supplementary budgets for the startup ecosystem. The J-StarX program run by JETRO, which I was selected for, is part of this initiative. The course targets pre-founding students with IT-related prototypes. Over three weeks from late August to mid-September 2024, I stayed on the Stanford University campus and participated in an intensive program led by USMAC (an accelerator) and Stanford.
Program Structure
Developing an entrepreneurial mindset; lectures on Silicon Valley's culture and uniqueness
Building business models and crafting pitches
Design Thinking bootcamp at Stanford's d.school, where Instagram Co-founder Mike Krieger once studied
Below is a summary of key takeaways.
Silicon Valley Culture
This quote captures the essence of Silicon Valley. Three layers sustain the region: the risk-taking culture forged during the Gold Rush, the technological foundation born from university and military research, and the talent drawn from around the world.
Because existing products fall short, founders build something better with their own hands. Google killed Yahoo; Uber killed the taxi. America's founding itself was a rejection of the old order ("We were the original Brexit"), and that spirit of rebellion lies at the heart of startup culture.
Defining Startups
A startup is an organization searching for a repeatable and scalable business model — in other words, a company where the product, the customer, and the monetization strategy are all still undefined. This definition reveals that startups are fundamentally "learning organizations."
An entrepreneur's daily routine revolves around the cycle of learn → innovate/adapt → teach (marketing). Understanding why customers buy and how they want to pay is the core of the job; technology is merely the means. The Week 1 session title, "It is not about your technology," expressed this philosophy directly.
Lessons from d.school
The Three P's: Perspectives, Portfolio, Process
The very first concept introduced was the three P's that run through all of d.school's teaching. Perspectives (incorporate diverse viewpoints), Portfolio (manage multiple ideas in parallel), and Process (distill into a repeatable method). Design thinking is not individual talent but a team-executable methodology — and d.school declares this upfront.
Exploration Funnel
At the core of d.school's curriculum is the Exploration Funnel framework. Starting from 4,000 ideas, you narrow down to 226 prototypes, then to 12 tested solutions, and ultimately arrive at 2–3 outcomes. This funnel visually demonstrates that design thinking is a process where quantity breeds quality.
Another critical framework is the Desirability / Viability / Feasibility Venn diagram. Is it desirable for users (Desirability)? Is it viable as a business (Viability)? Is it technically feasible (Feasibility)? Innovation emerges at the intersection of these three circles. What d.school consistently emphasizes is always starting from Desirability — what users actually want.
Double Diamond: Four Phases of Diverge and Converge
Day 1: Launch
The opening case study was GE's MRI redesign. Designer Doug Dietz witnessed a child crying in fear before an MRI scan. The conventional approach would be to improve the equipment, but instead he transformed the entire examination room into an "adventure journey." Scan completion rates and patient satisfaction improved dramatically. He didn't change the technology — he changed the experience.
Day 2, First Half: Diverge Problem
This phase is about expanding the problem space. The core method follows four steps: ENGAGE → NOTICE → FOLLOW-UP → SEEK STORIES.
First, you invite the user to talk: "Tell me about the last time you had that experience" (ENGAGE). Then you pick up on interesting observations (NOTICE), dig into the emotional layer with "How did that make you feel? Why?" (FOLLOW-UP), and finally identify the user's emotion and ask "Tell me about another experience where you felt the same way" (SEEK STORIES).
The purpose of this design is to elicit stories, not survey responses. People answer questionnaires rationally, but when they tell stories, their true feelings slip through.
Day 2, Second Half: Converge Problem
This phase derives insights from observations. The structure is simple but powerful.
"We met..." (who you spoke with) → "We were surprised they said..." (unexpected statements) → "We wonder if this means..." (what it implies). Here, four questions — "What does this person THINK?", "What do they FEEL?", "What do they BELIEVE?", "What are they MOTIVATED by?" — enable the "leap" from surface-level observation to the user's deeper psychology.
Next, you construct a problem statement: "It would be game changing to..." — and what you write here is the problem, not the solution. This constraint is critically important; it prevents premature solution-jumping and forces focus on defining the essential challenge.
Cautionary Tale: Preeti's Case
Preeti, a Stanford staff member, spent six months developing a solution before presenting it to anyone else for the first time. The result: "I realized others are not as eager as I am about the solution." By failing to validate the problem early, she wasted six months.
Day 2 Late – Day 3, First Half: Diverge Solution
This phase is about generating ideas at volume. The tool of choice is the "Yes, And!" Brainstorm.
First, convert the problem statement into a "How might we...?" question. Then each member individually proposes one idea, and the group expands it using the "Yes, And!" rule — no negation allowed.
d.school employs hybrid brainstorming (based on the Korde & Paulus study), cycling through individual ideation → group ideation → individual ideation. Group-only sessions fall prey to conformity pressure; solo-only sessions narrow the field of vision. Alternating between the two maximizes creativity.
Also notable is Justin Berg's Primal Mark research: the first idea proposed becomes an anchor that constrains subsequent thinking. To break this bias, deliberate constraints ("What if the budget were unlimited?", "What if we solved this with VR?", "What if it had to go viral on TikTok?") and analogies ("How would a sci-fi writer solve this?", "How would an ER doctor?") are introduced.
Ideas are evaluated along two axes: "highest likelihood of achieving the game changer" and "greatest value creation potential."
Day 3, Second Half – Day 4: Converge Solution
This phase validates selected ideas. The Describe Your Solution worksheet defines the product name, One Key Function, target customer, and intended impact.
The key frame introduced here is "We are trying to learn whether... this does that" — "We are trying to learn whether our One Key Function solves the Game Changing Problem." This positions the idea not as something to prove right, but as something to learn from. This is the philosophical core of d.school.
Recurring throughout the deck is the TEST YOUR ASSUMPTIONS cost curve. Error costs are low early in a project and grow exponentially over time. That's why Lo-Res Props — rough models made from pipe cleaners and paper — are used to validate hypotheses quickly. Speed takes priority over polish.
This warning about confirmation bias was repeated throughout the deck. To counteract the unconscious tendency to confirm hypotheses, you must actively seek disconfirmation. Test results are tracked using the Track Your Testing worksheet, recording per customer: "Does the customer want this Game Changer? (Yes/No)" and "Does the One Key Function achieve the intended impact? (Yes/No)." When No's accumulate, you rewrite the problem statement itself.
The final day's Design Review integrates the Problem side (user → surprising statement → deep motivation → Game Changing Problem) and Solution side (product name → key function → target → impact) into a single worksheet for cross-team feedback.
The Real Process Is Not Linear — The First Data Journey
The most instructive session on Day 4 was J. Scott Sanchez's practitioner report. In developing First Data's mobile wallet, the team initially defined "mobile payments" as the problem. But after observing a real user named Tanya and directly experiencing her behavior at drive-throughs and parking lots, they realized the true problem was "minimizing time spent waiting in line." They redefined the product around four functions — Pay/Earn/Redeem/Save — and iterated on a rough FAST food app prototype (5-screen mockup).
The First Data Journey diagram presented in this case was striking. Rather than a linear arrow from DP→CP→DS→CS, it depicted a nonlinear trajectory — looping through Fail, Reset, Backup, and Client Push Back, moving back and forth between phases repeatedly. The "process" of design thinking is not a textbook sequence of steps; in reality, it is a continuous, chaotic exploration.
Sanchez's Three Lessons
Start small — begin small and learn from setbacks
Go where there's suction — start with receptive people and places
Don't pick a side except the customer's — stand only on the customer's side
Conclusion
The most important principle I internalized at d.school is this: rather than crafting a perfect plan before acting, validate hypotheses with minimal prototypes and course-correct quickly. In the deck's own words, "DESIGNERS CREATE EXPERIENCES TO LEARN RAPIDLY." This is the practical embodiment of the same philosophy taught in Week 1 — the normalization of failure.