A rear bike light that sees cars coming before you do.
Risks: battery life with AI running, moulds and minimum orders, cash tied up in stock.
A lean product simulation
The 12-Week Accelerator
You get $500k and twelve weeks to take an AI startup from idea to validation. Spend wisely, talk to customers, and earn a term sheet at Demo Day.
Pick your startup
Four AI businesses, each with its own customers, risks and twelve weeks of decisions. Your first assumption about who the customer is will probably be wrong. That’s the point.
Snap your fridge. Get three dinners and a shopping list.
Risks: retention after the wow demo, inference costs, “ChatGPT already does this free”.
Get paid for the job without spending Friday on the phone.
Risks: trusting an agent with your customers, pilots that never convert, integrations.
Floors cleaned overnight by a robot, paid monthly.
Risks: robots needing rescue, hardware cost versus fees, safety around people.
How a week works
Every week you decide, act and live with the consequences. Matt, your Lean CPTO, gives you his verdict on every call, with the reasoning and the bill.
Plan the sprint
Spend three team slots: interviews, building, doing it by hand, growth tests, fixing tech debt or investor coffees.
Open the case file
Proto and Beta tell how real products won or lost: Palm, Zappos, Dropbox, Instagram, Pebble, Humane and more.
Make the call
Your startup’s real data, three options, and a price on each. Cash goes down. Evidence goes up, if you chose well.
Survive the curveball
A big tech rival launches. Your code breaks. An angel offers money with strings. Users churn. The burn comes out.
Demo Day
Hit four milestones (problem, solution, willingness to pay, retention) with cash still in the bank. Then Vale, the investor, reads your numbers.
Term sheetEvidence, milestones and runway.
Come back next quarterSomething real, not proven yet.
Vale passesLots built, too little learned.
Out of runwayExpensive calls made too early.
Twelve weeks, twelve lean skills
Each week teaches one move, grounded in a real case file, and ends with something you can use on Monday.
Week 1Test your riskiest assumptionPalm Pilot’s block of wood
Week 2Customer discoveryThe Mom Test
Week 3Fake it by hand firstZappos’ Wizard of Oz MVP
Week 4Smoke-test demandDropbox’s demo video
Week 5Scope to one jobBurbn becomes Instagram
Week 6Willingness to payPebble’s $10M Kickstarter
Week 7Actionable metricsIMVU and cohorts
Week 8One-way doorsHumane AI Pin
Week 9Pivot or persevereNintendo and the Wii
Week 10Product-market fitSuperhuman’s 40% test
Week 11Win a beachheadTesla’s Master Plan
Week 12Pitch the outcome“1,000 songs in your pocket”
Your guides
One real Lean CPTO, two veteran storytellers, a co-founder who wants fourteen features, and an investor who has heard every pitch.
The Lean CPTO
Matt
Matt Ainsworth helps founders build products for AI and physical AI without burning their runway on things nobody wants. In the game he frames every call in lean terms and tells you straight when it cost you.
Hardware stories
Proto
Eleven years as a test rig on a factory line. Has watched a lot of products go wrong in steel.
Software stories
Beta
Has been the beta of about forty startups. Knows the parts founders leave out of the keynote.
Your co-founder
Rex
Vibe coded the app in a weekend. Has a plan. It has a lot of AI features.
The investor
Vale
Has funded plenty of fourteen-feature plans. Would like to fund one that works.
Who it’s for
Aspiring product managers
Practise the calls PMs actually make: what to test, what to build, what to measure, and when to say no.
First-time founders
Feel how fast $500k goes, and why talking to customers beats shipping features, before it’s your own money.
Teams and cohorts
Running an accelerator, a product course or a team offsite? Talk to Matt about a facilitated session.
This week’s challenge
One startup, the same curveballs for everyone, scored out of 1000 on outcome, milestones, evidence, cash and lean calls. Here’s who’s on top.
The only pricing on this page
How many flat whites is this worth?
The accelerator is free. If it saved you from building something nobody wanted, you can place a bid in the only currency that matters.
3 flat whites
Place my bid: US$15
Completely optional, and a bit silly. Flat whites are US$5 each, roughly a Sydney flat white plus tip. You pay in US dollars through Stripe and can change the amount before you pay.
Demo Day is twelve weeks away
Pick a startup, plan your first sprint and find out whether you’d have earned the term sheet.
Start the accelerator