Founder/CEO @ LEANSTACK | Bestselling author and founder

I have bootstrapped several products to millions of users and worked with thousands of early-stage entrepreneurs worldwide - helping them systematically find repeatable and scalable business models using my battle-tested playbooks. 

Instead of:
-Spending several months building a product you hope people will buy.

Imagine a world where you:
- Spend several weeks defining a product you know people will buy.

For aspiring or early-stage founders:

If you’re looking to bootstart your idea, consider joining my 30-day Business Model Design Challenge below: 

You'll learn how to:
- Find or test a big idea worth pursuing,
- Design a business model blueprint,
- Stress-test your idea before investing time, money, and effort. 

Join the next Business Model Design Challenge:
https://runlean.ly/challenge

Ash Maurya

I've been working on something new.

After a decade of teaching Running Lean through frameworks and workshops, I realized something was missing. Founders understood the Lean Canvas intellectually. They could explain Demo-Sell-Build. They got it.

And then they'd go right back to building inside their heads.

The problem wasn't understanding. It was *feeling*. You have to feel the Innovator's Bias before you can overcome it.

So I did something I've never done before.

I wrote a novel.

THE INNOVATOR'S BIAS — a startup novel. Coming April 7th.

Pre-order now on Kindle.

I help founders build business models that scale. Author and creator of Lean Canvas.

These 2 boxes of the Lean Canvas are the outer pillars that carry your business model.

If you get these assumptions wrong, it’s easy to see how everything else in your business model falls apart.

- You end up describing a solution that no one wants (not desirable).
- Even if you build the solution you're already envisioning (feasible),
- No one buys it (not viable).

undefined undefined

The Just Start Manifesto (updated)
==========================
1. Entrepreneurs Are Everywhere.
2. There is No Better Time to Start.
3. Most Products Still Fail.
4. The Number One Reason Why Products Fail: Building Something Nobody Wants. 
5. The Number Two Reason Why Products Fail: Not starting. 
6. You Don’t Need Permission to Start.
7. Don’t Write a Business Plan. Sketch a Business Model.
8. Take Stock of Your Key Beliefs.
9. Take Stock of Your Key Metrics.
10. Take Stock of Your Key Insights.
11. Have Faith in Yourself. Be Ruthless With Your Ideas.
12. Don't Start With an MVP.
13. Not Faux Validation, But Traction
14. Love the Problem, Not Your Solution.
15. Sell Before You Build.
16. Bootstrap Your Way to Early Traction.
17. Focus on Time Versus Timing.
18. Value Your Time Correctly.
19. Avoid Premature Optimization.
20. Practice Right Action, Right Time.
21. Speed of Learning is the Only Unfair Advantage.
22. Speed of Learning x Speed of Execution = 100x Founder.
23. Remove Failure from Your Vocabulary.
24. It’s Time to Act on Your Big Idea.
25. Just Start.

I build systematic frameworks for early-stage founders. Author and creator of Lean Canvas.

The Just Start Manifesto

A feature isn't "done" when the code is shipped.

It's done when it provides validated learning from customers.

But validated learning is still a trailing indicator of what customers want. 

We can do better. 

Demand validation, sell before you build, is a leading indicator of what customers want.

This simple redefinition changes everything about your development process.

Demo-Sell-Build >> Build-Demo-Sell

The Weekend Canvas Exercise

Try this right now. It takes 30 seconds.

Pull up your Lean Canvas. Cover the Solution box with your hand.

Now explain why your business matters.

If you struggle — if the pitch falls apart without the solution — you're building solution-first.

The problem and customer should carry the pitch on their own. The solution is just the delivery mechanism.

I do this exercise every time I update my canvas. It's uncomfortable every single time. That discomfort is the point.

When I built CloudFire, the solution WAS the pitch. "A photo sharing app that..." Take away the app, and there was nothing. The problem wasn't specific enough to stand on its own.

When I built LeanSpark, the pitch without the solution was: "Founders waste months building products nobody wants because they skip structured discovery." That problem stands on its own. The solution is one of many possible responses.

Your problem should be so clear, so specific, and so painful that the audience nods before you ever mention what you built.

Try it. Let me know what you find.

Hot take: AI is making startups fail faster.

Not on purpose. But here's what's happening:

Founders use AI to generate code in hours. Ship MVPs in a weekend. Launch products in weeks instead of months.

Sounds great, right?

Except they're shipping untested ideas at record speed.

The build cost collapsed. But the failure rate didn't. Because the bottleneck was never "can I build it?"

It was always "should I build it?"

AI tools that help you build faster are accelerating the wrong part of the process. You're arriving at the wrong destination — just sooner.

The founders who win in the AI era aren't the ones who ship fastest. They're the ones who think clearest about what to ship.

That means starting with your business model. Stress-testing your assumptions. Running experiments before writing code.

LEANSpark is an AI tool that helps with the part most AI tools ignore — the thinking before the building.

Because the most expensive line of code is the one that solves a problem nobody has.

Free to try. No credit card required.

Every founder I coach thinks their business model is solid. Then we run the stress tests.

The pattern is always the same: strongest on Feasibility. Weak on everything else.

"Can I build it?" Yes.
"Do people want it?" I think so.
"Will the economics work at scale?" ...
"Why now and not 5 years ago?" ...
"What compounds in my favor over time?" ...

Two out of seven isn't a business model. It's a hunch with a prototype.

The blind spots don't kill you on day one. Cracks appear in month 9, and they kill you at month 18 when you've hired a team, raised money, and can't figure out why growth stalled.

What's the dimension you've never stress-tested?

I interviewed 47 customers. All of them said yes.

"Love the concept."
"Would definitely use this."
"When can I sign up?"

So I built it. 9 months. Full product. A photo-sharing app called CloudFire.

500 people signed up. Revenue trickled in. And then I did a 5-minute calculation that I should have done on day one, which proved my model wasn't viable for my business goal. 


47 people told me what I wanted to hear. Zero people told me the truth.

The lesson I carry everywhere now: Compliments are not evidence. Enthusiasm is not commitment. The only validation that matters is someone reaching for their wallet.

Three levels of customer evidence:
1. They SAY they have the problem
2. They SPEND money solving it today
3. They'll COMMIT money to YOUR solution

Most founders stop at level 1 and call it "validated."

Level 3 is where the truth lives.

What's the most misleading validation you've ever received?

Why "Talk to Customers" Is Bad Advice

"Talk to customers" is the most common startup advice. It's also the least useful.

Because HOW you talk to them matters more than WHETHER you talk to them.

Bad customer conversations:
- "Would you use a product that does X?" (Leading question — always gets yes)
- "What features would you want?" (They design your product for you — badly)
- "Is this a big problem?" (They agree to be polite)

Good customer conversations:
- "Walk me through the last time this happened." (Specific, past tense)
- "What did you do about it?" (Reveals actual behavior)
- "How much did that cost you?" (Quantifies the pain)

The difference: bad conversations confirm your bias. Good conversations reveal reality.

The single biggest upgrade is:

Stop asking what people would do. Start asking what they already did.

No alternative text description for this image

Meet the Lean(er) Canvas.

While sketching out a complete Lean Canvas is a great exercise for deconstructing an idea into a business model, here's an even faster ideation tool I use at the outset of all my ideas.

I focus attention on just the two outer boxes of the canvas: Customer segment and Problem.

These are the outer pillars that carry your business model.

If you get these assumptions wrong, it’s easy to see how everything else in your business model falls apart.

- You end up describing a solution that no one wants (not desirable).
- Even if you build the solution you're already envisioning (feasible),
- no one buys it (not viable).

Yes, just these 4 boxes are enough to prompt you to stress test your idea's desirability, viability, and feasibility.

If your idea doesn't pass these stress tests, your channels and unfair advantages won’t offer any respite.

Your business model is doomed.

Most Lean Canvases I see have this problem.

Trying to be two (or more) businesses at once.

I see this pattern constantly:

- Customer segment says "Small businesses AND enterprises"
- Revenue streams lists "subscriptions AND marketplace fees"
- Problem box describes two unrelated pain points

This is split model confusion. And it's the #1 reason for aimless wandering. 

Too many business models makes your canvas vague and you can't test a vague idea. 

The solution: create multiple smaller focused variants with exactly one business model on them. 

Here's how:

1. Detect the split axis
Is the tension in your customer segments? Your revenue model? Your channels? Your solution?

2. Stress-test independently
Run each variant through the same rigor: viability, desirability, and feasibility. The numbers will tell you which model has the most potential.

3. Choose with evidence, not intuition
Now you're not guessing which model to pursue. You're comparing validated data across focused alternatives.

This is the search mindset.

Most founders skip straight to execution with a single, broad model and spend months wondering why nothing clicks.

The best founders explore 2-3 focused models in parallel and converge on the winner fast.

I built this entire workflow into LEANSpark. It detects the split, generates focused variants in parallel, and lets you stress-test each one independently.

Watch it in action 👇

Simple models help us understand the complex. 

Before investing in any company, Warren Buffett and Charlie Munger use a latticework of mental models to evaluate the company. 

If it passes their tests, they move all in. Otherwise, they do nothing and wait.

Evaluating early-stage ideas is no different.

I have settled on 3 universal (product-agnostic) mental models that I use to
- baseline progress,
- prioritize risks, and
- formulate a “right action, right time” next step.

Here they are: 

1. Lean Canvas: Deconstruct an idea using 12 basic building blocks.

2. Traction Roadmap: Chart key milestones using 5 customer-centric metrics.  

3. Customer Forces: Understand customer behavior using 4 causal forces. 

BELIEFS + METRICS + INSIGHTS = BUSINESS MODEL STORY

Entrepreneurship is like a martial art — you can only learn by doing.

But "doing" without a sparring partner means burning months on the wrong thing.

LEANSpark is the sparring partner. Maps your business model, spots risky assumptions, and designs/runs experiments with you.

Free to try. No credit card.

Your Unique Value Proposition has 8 seconds to work.

That's how long first-time visitors spend on landing pages.

Forget complex benefit statements. Focus on the finished story benefit:

❌ "Eye-catching résumé templates"
✅ "Land your dream job"

Get inside your customer's head, not your product features.

The biggest waste in startups isn't money. It's effort spent on the wrong things at the wrong time.

I've worked with founders who worked 80-hour weeks and made zero progress — because they were doing Stage 3 activities at Stage 1.

Brand redesigns when they hadn't validated the problem. Growth hacking when they didn't have product-market fit. Hiring when they couldn't describe their customer.

Every stage has 2-3 actions that matter. Everything else is noise.

What stage are you at right now, and what are your 2-3?

#RightActionRightTime

Right Action, Right Time

We pre-sold 209 customers in 40 days.

Before writing a single line of production code.

Here's the Demo-Sell-Build playbook:

Most founders go Build > Demo > Sell. 

(Without vibe-coding):
Months of building. Then hoping someone buys.

(With Vibe-coding):
Days of building. Then months of trying to get someone to buy.

We flipped it: Demo > Sell > Build.

Step 1: DEMO.
We didn't build a product. We built a demonstration of the transformation. What does your life look like AFTER using this? We showed that. Perfectly okay to vibe-code here without going overboard on product. A demo is NOT meant to be used by customers. 

Step 2: SELL.
Not "would you use this?" — that's a survey.
Not "would you pay $50/mo?" — that's still a survey.
But "pay $150 to get early-access" — that's validation.
209 out of 307 people committed. 68% conversion rate.

Step 3: BUILD.
Now we knew exactly what to build, who it was for, and that they'd pay. No guessing. No hoping. Just executing against validated demand.

The fear: "But what if the demo isn't perfect?"
It doesn't have to be. That's the point. You're testing demand, not polish.

Building is the most expensive way to test an assumption. And yes, even when AI makes building cheaper, building without validation just digs a bigger future-sunk-cost fallacy sand trap. 

A demo costs hours. A product costs months.

The safest career move in 2026 isn't a promotion. It's validating a business model on the side.

Here's the math nobody talks about:

Your employer is actively figuring out how to do your job with AI and fewer people. That's not cynicism — that's their fiduciary duty to shareholders. Every earnings call now includes "AI efficiency gains." You're the efficiency being gained.

Meanwhile, the same AI that threatens your role has collapsed the cost of starting a company by 10x.

What used to require a co-founder, $500K in funding, and 18 months of runway — one person with domain expertise and AI can now validate in 90 days for under $1K.

The risk equation has flipped. But most people are still operating on the old math:
- Old math: Job = safe. Startup = risky.
- New math: Job = someone else controls your future. Validated startup = you control your future.

I'm not saying quit tomorrow. I'm saying: the skills you've built over 5, 10, 15 years in your industry? Those aren't just resume lines. They're founder superpowers.

You already know the problems. You already know the customers. You already know what's broken.

The only thing you haven't done is validate whether someone will pay you to fix it.

That's a weekend experiment, not a life decision.

What's the one problem in your industry that you see every day and nobody is solving well?

Here's a test for your "competitive advantage":

Can someone replicate it in 90 days with enough money?

If yes, it's not an advantage. It's a head start. And head starts disappear.

Things that are NOT advantages:
- "Our team" (good teams can be hired)
- "Our technology" (technology can be rebuilt)
- "First mover" (first movers get outcompeted by fast followers constantly)
- "Our features" (features get copied in weeks)

Things that ARE advantages:
- Network effects that compound (each user makes the product better for all users)
- Proprietary data that grows with usage (every interaction trains your system)
- Brand trust earned through years of consistent delivery
- Switching costs that deepen over time (not artificial lock-in — genuine integration depth)

The pattern: Real advantages get STRONGER the longer you operate. Fake advantages get WEAKER the more competitors enter.

Fill in the blank: "The longer we operate, _________ gets harder for competitors to replicate."

If you can't fill that in, your advantage is a head start, not a moat.

The Math That Changed How I Build

10 experiments at $100 each teach you more than 1 launch at $10,000.

Same total investment. 10x the learning.

The math:
- 10 experiments = 10 data points, 10 chances to be wrong, 10 opportunities to adjust
- 1 launch = 1 data point, 1 chance to be wrong, 1 expensive lesson

Early in my career, I optimized for launches. Big reveals. Grand openings. Each one felt significant. Most of them taught me one thing: I was wrong about something I could have tested for $100.

Now I optimize for learning velocity. How many assumptions can I test this month? Not how many features can I ship?

The founders who learn fastest don't have bigger budgets. They run smaller experiments.

What's one assumption you could test this week for under $100?

How to build your first startup moat.
Most founders don’t have an unfair advantage to start, and that’s okay.

A true unfair advantage is something that cannot be easily copied or bought. And finding one takes time.

The good news is that you don’t need one at the early stages, since all ideas sound a little crazy at first. 

So, embrace your obscurity as a gift and...

First, hunker down and focus on building a compelling Unique Value Proposition before your Unfair Advantage.

Then lean into what makes your compelling Unique Value Proposition truly unique — something different and better…

Turn that into a Counter-positioning strategy -- something incumbents can't copy without hurting their business model.

You’ll have the beginnings of your first unfair advantage.

Psychological injury claims do not appear out of nowhere.
They take shape when early warning signs are missed while work continues as normal.

I work with owners and senior leaders in desk-based SMEs (20–200 staff), where psychosocial risks tend to show up in meetings, workload decisions, and everyday digital communication.

I have seen this pattern repeatedly. A concern surfaces quietly. It is minimised, deferred, or absorbed because delivery pressure feels more urgent. The signal fades. Risk accumulates out of sight.

When it finally becomes visible, it is expensive and disruptive. A psychological injury claim. A formal complaint. A sudden exit. A manager collapse. Or a regulator or insurer conversation that nobody expected.

This is not a culture problem.
It is an unmanaged psychosocial risk problem. Psychosocial risk is managed through duty-holder clarity and consultation. It relies on living controls, not posters, slogans, or one-off initiatives.

After more than 20 years delivering large-scale technology and transformation programs, I kept encountering the same failure mode. Systems scale. Complexity increases. Signal flow breaks. When people stop speaking early, risk does not disappear. It goes underground.

I founded GRACEX to address that failure directly. Not by adding programs or surveys, but by helping catch psychosocial hazards early and deal with them while there is still room to act.

In practice, this means helping leaders establish clear ways to surface concerns, respond consistently, and follow through.
We do this through a simple early-signal loop: capture, triage, action, and close.

I believe working in a healthy workplace is a right, not a privilege. That belief is operational. When people have a safe way to raise issues early and see them handled properly, risk reduces, trust improves, and injury becomes less likely.

Psychosocial risk is not managed by intent or good values.
It is managed by whether problems can be raised early and resolved before they harden into injury claims.

Roland Illyes

𝗧𝗵𝗲 𝗰𝗼𝗿𝗽𝗼𝗿𝗮𝘁𝗲 𝗛𝗥 𝗽𝗵𝗿𝗮𝘀𝗲 𝗴𝘂𝗶𝗱𝗲 𝗻𝗼𝗯𝗼𝗱𝘆 𝗮𝘀𝗸𝗲𝗱 𝗳𝗼𝗿

I have spent this week being the bearer of alarming news. 
So today, a Friday gift: the ultimate corporate phrase guide:
"𝗪𝗲 𝗮𝗿𝗲 𝗮 𝗽𝗲𝗼𝗽𝗹𝗲 𝗳𝗶𝗿𝘀𝘁 𝗼𝗿𝗴𝗮𝗻𝗶𝘀𝗮𝘁𝗶𝗼𝗻." 𝘔𝘦𝘢𝘯𝘴: Company first. People second, when budget allows. 
"𝗣𝗲𝗼𝗽𝗹𝗲 𝗮𝗻𝗱 𝗖𝘂𝗹𝘁𝘂𝗿𝗲." 𝘔𝘦𝘢𝘯𝘴: HR got rebranded. Same corporate risk management. New lanyard. 
"𝗪𝗲 𝗵𝗮𝘃𝗲 𝗮𝗻 𝗼𝗽𝗲𝗻 𝗱𝗼𝗼𝗿 𝗽𝗼𝗹𝗶𝗰𝘆." 𝘔𝘦𝘢𝘯𝘴: The door is open. Nobody is stopping you from leaving. 
"𝗪𝗲 𝗮𝗿𝗲 𝗹𝗶𝗸𝗲 𝗮 𝗳𝗮𝗺𝗶𝗹𝘆 𝗵𝗲𝗿𝗲." 𝘔𝘦𝘢𝘯𝘴: Red tape, hierarchy, and nobody can explain who made that decision or why. 
"𝗧𝗵𝗲 𝗯𝘂𝘀𝗶𝗻𝗲𝘀𝘀 𝗵𝗮𝘀 𝗱𝗲𝗰𝗶𝗱𝗲𝗱..." 𝘔𝘦𝘢𝘯𝘴: Someone had an idea. Nobody knows whose. Nobody dared ask. Everybody is responsible for it. 
"𝗪𝗲 𝗿𝘂𝗻 𝗮𝗻 𝗮𝗻𝗻𝘂𝗮𝗹 𝗲𝗻𝗴𝗮𝗴𝗲𝗺𝗲𝗻𝘁 𝘀𝘂𝗿𝘃𝗲𝘆." 𝘔𝘦𝘢𝘯𝘴:  We measure how disengaged people are, once a year, after it stopped mattering. 
"𝗦𝗼𝗳𝘁 𝘀𝗸𝗶𝗹𝗹𝘀." 𝘔𝘦𝘢𝘯𝘴: Being human at work. Apparently optional.
Vocabulary note: 𝘛𝘩𝘦 𝘢𝘣𝘪𝘭𝘪𝘵𝘺 𝘵𝘰 𝘩𝘢𝘷𝘦 𝘢 𝘥𝘪𝘧𝘧𝘪𝘤𝘶𝘭𝘵 𝘤𝘰𝘯𝘷𝘦𝘳𝘴𝘢𝘵𝘪𝘰𝘯 𝘣𝘦𝘧𝘰𝘳𝘦 𝘪𝘵 𝘣𝘦𝘤𝘰𝘮𝘦𝘴 𝘢 𝘨𝘳𝘪𝘦𝘷𝘢𝘯𝘤𝘦. 𝘛𝘩𝘦 𝘢𝘣𝘪𝘭𝘪𝘵𝘺 𝘵𝘰 𝘴𝘱𝘰𝘵 𝘸𝘩𝘦𝘯 𝘴𝘰𝘮𝘦𝘰𝘯𝘦 𝘪𝘴 𝘴𝘵𝘳𝘶𝘨𝘨𝘭𝘪𝘯𝘨 𝘣𝘦𝘧𝘰𝘳𝘦 𝘵𝘩𝘦𝘺 𝘨𝘰 𝘰𝘯 𝘴𝘵𝘳𝘦𝘴𝘴 𝘭𝘦𝘢𝘷𝘦. 𝘛𝘩𝘦 𝘢𝘣𝘪𝘭𝘪𝘵𝘺 𝘵𝘰 𝘣𝘶𝘪𝘭𝘥 𝘢 𝘵𝘦𝘢𝘮 𝘵𝘩𝘢𝘵 𝘥𝘰𝘦𝘴𝘯'𝘵 𝘲𝘶𝘪𝘦𝘵𝘭𝘺 𝘧𝘢𝘭𝘭 𝘢𝘱𝘢𝘳𝘵.
𝘞𝘦 𝘤𝘢𝘭𝘭𝘦𝘥 𝘵𝘩𝘢𝘵 𝙨𝙤𝙛𝙩.

Happy Friday. 👋 
If any of these made you wince, you have probably sat in the same meetings I have.

Turning invisible office safety risks into accountable action in Aussie 20-200 staff businesses | Founder, GRACEX

160% increase would be impressive in an investment portfolio.

In serious workplace mental health claims, it is something else.
Safe Work Australia data shows that serious mental health claims have risen by about 160% over the past decade.

The median time off work for one of these claims is around 9 months.

In a 20 employee business, that is 5% of your workforce gone for most of a year.
In a 50 employee business, it is 2%.

You either hire temporary cover you stretch the team and increase pressure. One is more expensive than the other.

This is operational risk, not abstract policy debate.

Where is your early warning signal?

Line chart from Safe Work Australia showing an index of serious workers compensation claims by nature of injury from 2013–14 to 2023–24, with 2013–14 set to 100. The line for mental health conditions rises steadily to around 260 by 2023–24, while other categories such as musculoskeletal and connective tissue diseases, traumatic joint, ligament and muscle injuries, and wounds or lacerations increase more moderately to around 120–140 over the same period.

A team member resigned three weeks ago. The offboarding conversation was respectful, even warm. They said all the right things and left on good terms.

Three weeks later, a work stress claim arrived.

When the insurer asked for documentation, you could walk them through everything. A manager had flagged unrealistic timelines. The employee had raised that they were struggling. A conversation happened, deadlines were adjusted, and everyone got back to work.

You remembered it clearly. None of it was on paper.

No record of what was identified, what decision was made, or who owned the follow-up. Just a clear memory of genuinely good intentions and a process that felt handled at the time.

The dashboard had been green the entire time.
The gap was never awareness. It was the system between awareness and evidence.

𝘏𝘰𝘸 𝘮𝘶𝘤𝘩 𝘰𝘧 𝘺𝘰𝘶𝘳 𝘳𝘪𝘴𝘬 𝘮𝘢𝘯𝘢𝘨𝘦𝘮𝘦𝘯𝘵 𝘭𝘪𝘷𝘦𝘴 𝘪𝘯 𝘱𝘦𝘰𝘱𝘭𝘦'𝘴 𝘮𝘦𝘮𝘰𝘳𝘺?

Two-panel graphic. Left panel: "What you saw" — five green indicators showing a business that appeared healthy: no dashboard alerts, EAP in place, offboarding handled, team delivering, manager who flagged and resolved. Right panel: "What was building" — five red indicators showing what was invisible: undocumented conversations, no accountability trail, workload quietly absorbing the pressure, risk not visible in any system, a claim that arrived three weeks later.

taplio

The Battle

Ash Maurya

Roland Illyes

Followers

Posts

Engagements

Likes

Performance ScoreAsh Maurya's LinkedIn Performance

Performance ScoreRoland Illyes's LinkedIn Performance

Ready to grow your LinkedIn presence?

taplio

The Battle

Ash Maurya

Roland Illyes

Followers

Posts

Engagements

Likes

Performance Score.css-goub6s{font-size:var(--chakra-fontSizes-md);font-weight:500;color:var(--chakra-colors-gray-600);display:block;margin-top:var(--chakra-space-1);}Ash Maurya's LinkedIn Performance

Performance ScoreRoland Illyes's LinkedIn Performance

Ready to grow your LinkedIn presence?

Performance ScoreAsh Maurya's LinkedIn Performance