Startable

The situation

Launchpad exists to make new business creation happen inside Fujitsu. Ventures are one half. The other is people: employees who can think with an innovation mindset and build something, whether or not a venture comes out at the end.

That half was taught in a room. Venture associates on a six-month program, innovation teams elsewhere in the company through one-off seminars. A learning team of three ran all of it.

Three people can't reach a corporation one room at a time, and what anyone learned depended on who was free to teach that week.

Online was the obvious move, and the obvious version of online is the one that kills it. Record the seminars, add a quiz, call it a curriculum. This isn't information you absorb. It's a handful of habits you only own after using them badly in front of someone who'll say so. The brief, then: get the scale, keep the room.

We used the studio's playbook of sprints, validation and shipping small, minus the MVP posture. Will anyone want this was already answered. The seminars existed, the audience was in the building. The open question was whether learning survives the move online, and a thin slice can't test that. What's under test is whether a whole cohort experience holds.

Learning by doing

We started in the room, watching the learning team teach. What happens in an exercise, and what it's for. Then we went through them together, asking what the online version is: the same thing, a substitute, or nothing worth keeping.

I designed the blocks and their interactions from that, the creators reviewed them against their own material, and I onboarded the team onto the creator side. They've run it since. I still get asked for visual help. I don't get asked to build courses.

Reading and video are in the set, but they're setup. Everything else asks the learner to make something. There isn't one quiz, because the interaction is the exercise. Ranking forces a trade-off you can't hedge. Matching attaches a pattern to a company you know. Fill-in-the-blank hands you the studio's problem-statement format and makes you fill it with your own venture.

The heavier blocks are the room's exercises, shrunk. Rapid ideation is a sticky-note board with a ten-minute timer and an instruction not to self-edit. Five Whys is five fields and a rule against stopping at three.

The room, rebuilt

An exercise nobody sees is homework. The same exercise, seen by the people you'll be building next to, is practice.

Every activity carries its own comment column, with peers and facilitators writing in the same place and facilitators badged. Visibility is a setting, not an assumption: each piece of work has a public switch, because the fastest way to kill honest work is to make people perform.

The work accumulates into a portfolio grouped by journey. Not a completion certificate but a body of work, readable by the cohort. Facilitators get the same object from the other side: who's on which journey, who's finished what, one click into anyone's portfolio.

A kit, not a course

A platform three people can't publish to isn't infrastructure.

One content model: a journey holds stages, stages hold modules, modules hold sections, and a section is built from blocks. A six-month program and a ninety-minute seminar are the same object at different scales. Illustration, layout and states arrive with the blocks, so a course looks finished the day it's written.

Two languages, both directions. Almost all the good material on business creation is written overseas and the people who need it are here, so the language switch sits in the editor at section level and creators write in whichever language they're better in.

The harder gap isn't language. Vocabulary translates; a mindset doesn't. We localized rather than translated, using Japanese examples where they exist and overseas cases where they don't.

One small feature did the most work. Innovation vocabulary stays English even in Japanese practice, so every term links to a definition, an explanation and a case, and every term a learner meets collects into a glossary. It became one of the most-used pages in the product.

The visual direction came from the same idea, worked out with our brand designer and the venture lead: venture building as block building. A journey's roadmap is a building, and finishing a stage adds a floor. The gamification runs on that metaphor instead of beside it.

Whether it worked

Course satisfaction is the easy number and the least useful one. People enjoy things that teach them nothing.

So measurement went in early. Feedback at the activity rather than at the end, facilitator and peer responses as signal, analytics on where people slow or stop, interviews after each cohort. Not for a dashboard, but for knowing which activity to cut before the next cohort reached it.

Startable shipped v0 and v1, with fifty-odd learners through v1. Every venture associate is now onboarded through it, other Fujitsu innovation units run their learning on it, and it's still expanding inside the parent company. It carries one of the two things Launchpad exists to do.

I handed the product to another designer for v2 and stayed on the strategy above it.