From first idea
to repeat customers
AI builds now. The founder’s job is judgment.
Coaching and startup programs that train it, built on Design Thinking × AI-native Venture Building.
In the last four weeks,
how many customers did you talk to?
Building got easy.
Selling didn’t.
Across the teams coached, the common weakness wasn’t the product.
It was finding customers, selling, and getting to market.
Plenty built. Nothing ready to sell.
One team had closed 72 of 78 development tickets.
The price was still undecided.
What gets called the bottleneck usually isn’t one
Six teams were asked what was most urgent: developers, funding, payments, legal.
Not one said “finding out if customers will really pay to use it.”
What AI can’t replace is the years a founder has put in
A founder with over a decade in one field left it off page one of the deck.
So did a founder with a PhD, and one with more than twenty years in a single industry.
Only after someone outside pointed at it did all three say, “that was our edge.”
Questions from outside surface the Unfair Advantage you can’t see from inside.
The founders we grow
From function generalist
to problem specialist
Product Builder and Business Builder
Founders who decide what to make, who it’s for, and how it sells,
and own it to the end.
AI brought the cost of how to build close to zero.
So the competition moved to what, for whom, and how it sells.
Using the tools well is now only the starting line.
Why principles matter more now
| What changed | What happens | What matters more |
|---|---|---|
| Building costs near zero | Teams skip testing and build the wrong thing faster | Problem definition |
| Supply explodes | Products flood in, and customer attention gets scarce | Distribution · Customer acquisition |
| Copying costs near zero | Features get matched within a day | Brand · Trust |
Seven principles
| Principle | The question |
|---|---|
| Problem | Is it a problem people would pay to solve? |
| Customer | Who uses it, and who pays? Who are the first 100? |
| Value | With alternatives out there, why this one? |
| Distribution ★ | Where do you meet customers, and what does one cost to bring in? |
| Revenue | Who pays, how much, and why? Is anything left over? |
| Scale | Does the same approach repeat from one customer to ten thousand? |
| Defense ★ | If someone rebuilds it with AI, what’s left? |
★ The two that changed most after AI.
The growth path — how many questions you own
| Stage | Principles owned | What it leaves behind |
|---|---|---|
| Specialist | One or two, in your own field | Screens, code, papers, contracts |
| Product Builder | Problem · Customer · Value | A product people use |
| Business Builder | + Distribution · Revenue · Scale · Defense | Sales |
Everyone starts somewhere different
Designer
- Have
- Problem · Customer · Value · Defense
- Missing
- Distribution · Revenue · Scale
Easy to feel done once the screens are done.
Developer
- Have
- Value · Scale
- Missing
- Problem · Customer · Distribution · Revenue · Defense
Easy to believe the features are the moat.
Researcher
- Have
- Problem · Defense
- Missing
- Customer · Distribution · Revenue · Scale
Knows the problem best, and tends to meet the person who pays last.
Sales background
- Have
- Customer · Distribution · Revenue
- Missing
- Problem · Value · Scale
Can sell first, but the reason it sold may not be the product.
Wherever you start, filling in the missing principles is the path to Business Builder.
Business principles aren’t a field to cover.
They’re the minimum grammar for owning one problem to the end.
We don’t give answers.
We change the question.
Five stages, from one line of an idea to a customer who buys again.
Each opens with one question.
Discover · Problem, Customer
Have you met the customer written into your grant application?
Meet real customers instead of the ones on paper, and let AI organize what they said.
Define · Problem, Value
Is the story you tell investors the one you tell customers?
Pull the scattered stories into one sentence for the business, and set the rest down.
Build · Value — AI executes, the founder judges
Three days to build it. Why still explain it in words?
The founder builds the prototype with AI. Only what you built yourself can be fixed the next morning.
Test · Distribution
How many chances did you give customers to say no this week?
It shows where customers are, and what one costs to reach.
Pay · Revenue, Scale, Defense
Do you know why your first buyers didn’t buy again?
The process runs past the test, where design thinking stops, all the way to the repeat purchase.
A customer who comes back is trust, and trust is the defense AI can’t copy.
It starts from design thinking’s discovery process, then changes where it ends, when you build, and who builds.
When building took six months, polishing was care.
Now that it takes three days, it’s avoidance.
Designing Agentic Experiences
Three ways in
What’s left is one thing rebuilt eight times. Details → One semester Teams get diagnosed and sorted into tracks in week one.
The same five stages, at a different speed per track. Details → Three days Build and test, compressed into three days.
What’s left is three rejections. Details →
Start here
Apply, and you’ll get a proposal built around where you are.
For institutions and operators, we propose course structure and how it would run.
For founders, format, schedule, and fee.
Nothing needs to be sorted out yet. One line on what you’re working on is enough.
Experience
- Corporate venture officeRan internal ventures inside a company
- Scaling a businessFrom 100 to 1,000 inside a company
- Founded a startupBuilt it from 0 to 1
- Government grant judgeEvaluating startup grant applications
- Startup programsMentor · Lecturer · Judge
Experience from 0 to 1 and from 100 to 1,000, put to work beside founders.
The full course,
or just the stages you need
A · Institutions — one-semester course
One semester. Teams, sorted.
- Course title
- Design Thinking × AI-native Venture Building
- Course goal
- Problem specialists — Product Builders and Business Builders who own all seven principles
Run the same curriculum across eight teams and two of them move.
A team with paying customers and a team without one have different work next week.
Teams get sorted in week 1, execution gets logged every week,
and the final week is measured against week 1, in numbers.
In the same pre-launch cohort, a team shipping its own MVP with vibe coding
sits next to a team that opened GPT for the first time this month.
So week 1 sorts that gap first.
Otherwise the team behind never catches up, and the team ahead gets bored.
What you keep
| Document | When |
|---|---|
| Diagnostic sheet | Week 1 · the basis for each team’s track |
| Weekly execution log | Every week · done, stopped, rejected |
| Prototype and revision history | Every week · what changed and why |
| Closing re-measurement report | Final week · change against week 1, in numbers |
B · Program operators
Just the stages you need
The operator designs and runs the program.
Inside it, gain ground takes the stages that get teams actually moving.
| Module | What it covers |
|---|---|
| Week-one diagnostic workshop | Through each team’s diagnostic sheet, the basis for splitting tracks |
| Curriculum stage modules | Only the stages the program needs: Discover, Define, Build, Test, Pay |
| 1:1 coaching track | Weekly coaching per team for the length of the program (eight-week structure) |
| Demo day prep clinic | Fitting what you tell investors and what you tell customers into one sentence |
| Three-day intensive | Build and test compressed, in a bootcamp or hackathon format |