//
Your MVP Is a Graveyard
Most founders bury their product before it ever ships. Here's how we keep them above ground.
//Enterprise Software
[overview]
title:
Your MVP Is a Graveyard
date:
[topics]
Product Strategy
Founders
Product Market Fit
The 47-Feature MVP
The Notion doc was forty-seven rows long.
The founder slid it across the table, toggles, dashboards, a billing portal, a "social tab," dark mode, three onboarding flows, a referral engine, an admin panel nobody had defined. Every row had a checkbox. Every checkbox was unticked. He called it the "MVP scope."
He had six weeks of runway, one developer he hadn't hired yet, and a wife asking him when the thing was going to make money.
We didn't argue. We just turned the doc around and started crossing things out with a red pen. By the time we got to row twenty-three he stopped us. "Wait, if you cut all of that, what's even left?"
That's the question. That's always the question. And the answer is the only thing that ever mattered.
TL;DR, If your "MVP" has more than ten features, you're not building a product, you're digging a graveyard. The fix isn't more discipline, it's a smaller, sharper bet: one feature people would scream about losing, conversations that don't mention your solution, and a smoke test that costs less than your domain renewal. Speed is the strategy. Done is the teacher.
The Problem: Founders Confuse Completeness With Confidence
Every founder we meet has the same hallucination. A perfect launch day. A polished platform. Users marvelling at the settings menu. Press. A graph that goes up and to the right.
What they actually have is a backlog. A long, beautifully-organised, completely unvalidated backlog. The features feel like progress because they look like a product. But every row in that doc is an unverified assumption dressed up as a deliverable.
Here's the failure mode, named plainly: founders build to feel safe, not to learn fast. A 47-row MVP isn't ambition, it's avoidance. It's the founder buying themselves three more months of "we're not ready to launch yet" before the market has a chance to tell them they're wrong.
The cost isn't just time. It's worse. Every extra feature dilutes the one thing the product is supposed to be unmistakably good at. By the time you ship, your core value is buried under a mountain of nice-to-haves. The user doesn't know what you're for. Neither do you anymore.
The agencies don't help, they quote against the 47-row doc, because that's how they bill. The no-code influencers don't help, they'll happily sell you the stack to build all of it. And the founder community doesn't help, because everyone's quietly comparing scope and nobody wants to be the one who shipped "less."
The graveyard is full of products that were almost ready.
How We Do It: The Five Cuts We Make Before A Line of Code Ships
When that founder asked what was even left after we crossed out twenty-three rows, we showed him the actual MVP framework we use at the studio. Five cuts. In this order. No skipping.
1. The Photo Filter Test
Instagram didn't launch as a social network. It launched as a filter app. No comments. No stories. No reels. One job: make a phone photo look better. They added everything else after people would have walked into traffic to keep using it.
So we ask every founder: what's your photo filter? The one thing your user would feel physically pained to lose? If you can't answer in five seconds, you're about to spend six months building features nobody asked for.
For the 47-row founder, the answer turned out to be a single button, "price my project." Forty-six other features were scaffolding for that one moment.
2. The Un-Interview
Most founder "user research" is a confession booth. They show a prototype, ask "would you use this?", and collect the polite yeses they came for. It's a waste of 20 calls.
We run un-interviews. The rule: you never mention your solution. You ask the user to describe the last time the problem made them want to throw their laptop out the window. You're not a salesperson, you're an investigator. They are symptom reporters, not solution designers. The Voice of the Customer is not an instruction manual. It's a map of pain.
If you can't write your user's headache in one sentence after twenty calls, you haven't listened long enough.
3. The Ethical Smoke Test
Demand is a hallucination until someone gives you money, time, or a real intent signal. Before we touch a repo, we put up a smoke test landing page, the product written as if it already exists, with a real promise and a real price.
When a user clicks Buy, they get the truth: "We're not live yet. Want to be first in?" No tricks. No bait. Just a clean signal.
The numbers tell you what to do next:
3–5% conversion — weak. Pivot or re-frame.
10–20% conversion — green light. Build the photo filter.
>20% conversion — warning bell. The pull is so strong that if your backend is fragile, it'll catch fire on day one.
That last one is the trap we see most often. Founders treat virality as a finish line. It's an ignition source.
4. The Rebuild-Ready Foundation
Most MVPs don't die from lack of features. They die the week traction finally arrives, when the architecture chosen for "just validating" collapses under the first real load. We've inherited too many of those repos to think it's bad luck.
So we tell every founder the same thing: ship fast, but architect like you'll be glad you did in month six. A real relational data model from day one. Clear boundaries between the product and the integrations bolted onto it. Automated tests for the one workflow that actually matters, the photo filter from cut #1. CI/CD before you have anyone to deploy for.
This isn't gold-plating, and it isn't premature optimisation. It's the difference between a launch that compounds and a launch that has to be rebuilt mid-fire. Rebuild-ready isn't pessimism. It's the only posture that survives the moment your smoke test was right.
5. The Disappointment Metric
Sign-ups are a vanity metric. So is GMV in week one. So is press. The only PMF question that matters is the Sean Ellis test:
How would you feel if you could no longer use this product?
If 40% or more of your active users say "very disappointed," you've hit Product-Market Fit. If they shrug, you haven't — no matter how good the chart looks. PMF isn't a switch. It's a spectrum, and it moves. A competitor ships a delighter, expectations reset, and your "fit" decays. That's why retention curves matter more than acquisition charts. Flat retention is the only honest signal.
Conclusion: The Only Button That Matters
The 47-feature founder shipped six weeks later. One screen. One button. Price my project. Forty-three of his original rows never got built — and never needed to. The product is live. He's paying himself. His wife stopped asking.
An MVP isn't a destination. It's an experiment whose only job is to earn you a second date with your user. Speed of execution is your only real competitive advantage. Perfectionism is just a delay tactic for founders afraid to be wrong in public.
So here's the question we leave every founder with, the same one we asked across that table, six weeks before the thing went live:
If you had to launch tomorrow with one button, what would that button do?
Go build that button. Done is the only thing that will ever teach you how to be perfect.
//
Your MVP Is a Graveyard
Most founders bury their product before it ever ships. Here's how we keep them above ground.
//Enterprise Software
[overview]
The 47-Feature MVP
The Notion doc was forty-seven rows long.
The founder slid it across the table, toggles, dashboards, a billing portal, a "social tab," dark mode, three onboarding flows, a referral engine, an admin panel nobody had defined. Every row had a checkbox. Every checkbox was unticked. He called it the "MVP scope."
He had six weeks of runway, one developer he hadn't hired yet, and a wife asking him when the thing was going to make money.
We didn't argue. We just turned the doc around and started crossing things out with a red pen. By the time we got to row twenty-three he stopped us. "Wait, if you cut all of that, what's even left?"
That's the question. That's always the question. And the answer is the only thing that ever mattered.
TL;DR, If your "MVP" has more than ten features, you're not building a product, you're digging a graveyard. The fix isn't more discipline, it's a smaller, sharper bet: one feature people would scream about losing, conversations that don't mention your solution, and a smoke test that costs less than your domain renewal. Speed is the strategy. Done is the teacher.
The Problem: Founders Confuse Completeness With Confidence
Every founder we meet has the same hallucination. A perfect launch day. A polished platform. Users marvelling at the settings menu. Press. A graph that goes up and to the right.
What they actually have is a backlog. A long, beautifully-organised, completely unvalidated backlog. The features feel like progress because they look like a product. But every row in that doc is an unverified assumption dressed up as a deliverable.
Here's the failure mode, named plainly: founders build to feel safe, not to learn fast. A 47-row MVP isn't ambition, it's avoidance. It's the founder buying themselves three more months of "we're not ready to launch yet" before the market has a chance to tell them they're wrong.
The cost isn't just time. It's worse. Every extra feature dilutes the one thing the product is supposed to be unmistakably good at. By the time you ship, your core value is buried under a mountain of nice-to-haves. The user doesn't know what you're for. Neither do you anymore.
The agencies don't help, they quote against the 47-row doc, because that's how they bill. The no-code influencers don't help, they'll happily sell you the stack to build all of it. And the founder community doesn't help, because everyone's quietly comparing scope and nobody wants to be the one who shipped "less."
The graveyard is full of products that were almost ready.
How We Do It: The Five Cuts We Make Before A Line of Code Ships
When that founder asked what was even left after we crossed out twenty-three rows, we showed him the actual MVP framework we use at the studio. Five cuts. In this order. No skipping.
1. The Photo Filter Test
Instagram didn't launch as a social network. It launched as a filter app. No comments. No stories. No reels. One job: make a phone photo look better. They added everything else after people would have walked into traffic to keep using it.
So we ask every founder: what's your photo filter? The one thing your user would feel physically pained to lose? If you can't answer in five seconds, you're about to spend six months building features nobody asked for.
For the 47-row founder, the answer turned out to be a single button, "price my project." Forty-six other features were scaffolding for that one moment.
2. The Un-Interview
Most founder "user research" is a confession booth. They show a prototype, ask "would you use this?", and collect the polite yeses they came for. It's a waste of 20 calls.
We run un-interviews. The rule: you never mention your solution. You ask the user to describe the last time the problem made them want to throw their laptop out the window. You're not a salesperson, you're an investigator. They are symptom reporters, not solution designers. The Voice of the Customer is not an instruction manual. It's a map of pain.
If you can't write your user's headache in one sentence after twenty calls, you haven't listened long enough.
3. The Ethical Smoke Test
Demand is a hallucination until someone gives you money, time, or a real intent signal. Before we touch a repo, we put up a smoke test landing page, the product written as if it already exists, with a real promise and a real price.
When a user clicks Buy, they get the truth: "We're not live yet. Want to be first in?" No tricks. No bait. Just a clean signal.
The numbers tell you what to do next:
3–5% conversion — weak. Pivot or re-frame.
10–20% conversion — green light. Build the photo filter.
>20% conversion — warning bell. The pull is so strong that if your backend is fragile, it'll catch fire on day one.
That last one is the trap we see most often. Founders treat virality as a finish line. It's an ignition source.
4. The Rebuild-Ready Foundation
Most MVPs don't die from lack of features. They die the week traction finally arrives, when the architecture chosen for "just validating" collapses under the first real load. We've inherited too many of those repos to think it's bad luck.
So we tell every founder the same thing: ship fast, but architect like you'll be glad you did in month six. A real relational data model from day one. Clear boundaries between the product and the integrations bolted onto it. Automated tests for the one workflow that actually matters, the photo filter from cut #1. CI/CD before you have anyone to deploy for.
This isn't gold-plating, and it isn't premature optimisation. It's the difference between a launch that compounds and a launch that has to be rebuilt mid-fire. Rebuild-ready isn't pessimism. It's the only posture that survives the moment your smoke test was right.
5. The Disappointment Metric
Sign-ups are a vanity metric. So is GMV in week one. So is press. The only PMF question that matters is the Sean Ellis test:
How would you feel if you could no longer use this product?
If 40% or more of your active users say "very disappointed," you've hit Product-Market Fit. If they shrug, you haven't — no matter how good the chart looks. PMF isn't a switch. It's a spectrum, and it moves. A competitor ships a delighter, expectations reset, and your "fit" decays. That's why retention curves matter more than acquisition charts. Flat retention is the only honest signal.
Conclusion: The Only Button That Matters
The 47-feature founder shipped six weeks later. One screen. One button. Price my project. Forty-three of his original rows never got built — and never needed to. The product is live. He's paying himself. His wife stopped asking.
An MVP isn't a destination. It's an experiment whose only job is to earn you a second date with your user. Speed of execution is your only real competitive advantage. Perfectionism is just a delay tactic for founders afraid to be wrong in public.
So here's the question we leave every founder with, the same one we asked across that table, six weeks before the thing went live:
If you had to launch tomorrow with one button, what would that button do?
Go build that button. Done is the only thing that will ever teach you how to be perfect.
title:
Your MVP Is a Graveyard
date:
[topics]
Product Strategy
Founders
Product Market Fit
//
Your MVP Is a Graveyard
Most founders bury their product before it ever ships. Here's how we keep them above ground.
//Enterprise Software
[overview]
The 47-Feature MVP
The Notion doc was forty-seven rows long.
The founder slid it across the table, toggles, dashboards, a billing portal, a "social tab," dark mode, three onboarding flows, a referral engine, an admin panel nobody had defined. Every row had a checkbox. Every checkbox was unticked. He called it the "MVP scope."
He had six weeks of runway, one developer he hadn't hired yet, and a wife asking him when the thing was going to make money.
We didn't argue. We just turned the doc around and started crossing things out with a red pen. By the time we got to row twenty-three he stopped us. "Wait, if you cut all of that, what's even left?"
That's the question. That's always the question. And the answer is the only thing that ever mattered.
TL;DR, If your "MVP" has more than ten features, you're not building a product, you're digging a graveyard. The fix isn't more discipline, it's a smaller, sharper bet: one feature people would scream about losing, conversations that don't mention your solution, and a smoke test that costs less than your domain renewal. Speed is the strategy. Done is the teacher.
The Problem: Founders Confuse Completeness With Confidence
Every founder we meet has the same hallucination. A perfect launch day. A polished platform. Users marvelling at the settings menu. Press. A graph that goes up and to the right.
What they actually have is a backlog. A long, beautifully-organised, completely unvalidated backlog. The features feel like progress because they look like a product. But every row in that doc is an unverified assumption dressed up as a deliverable.
Here's the failure mode, named plainly: founders build to feel safe, not to learn fast. A 47-row MVP isn't ambition, it's avoidance. It's the founder buying themselves three more months of "we're not ready to launch yet" before the market has a chance to tell them they're wrong.
The cost isn't just time. It's worse. Every extra feature dilutes the one thing the product is supposed to be unmistakably good at. By the time you ship, your core value is buried under a mountain of nice-to-haves. The user doesn't know what you're for. Neither do you anymore.
The agencies don't help, they quote against the 47-row doc, because that's how they bill. The no-code influencers don't help, they'll happily sell you the stack to build all of it. And the founder community doesn't help, because everyone's quietly comparing scope and nobody wants to be the one who shipped "less."
The graveyard is full of products that were almost ready.
How We Do It: The Five Cuts We Make Before A Line of Code Ships
When that founder asked what was even left after we crossed out twenty-three rows, we showed him the actual MVP framework we use at the studio. Five cuts. In this order. No skipping.
1. The Photo Filter Test
Instagram didn't launch as a social network. It launched as a filter app. No comments. No stories. No reels. One job: make a phone photo look better. They added everything else after people would have walked into traffic to keep using it.
So we ask every founder: what's your photo filter? The one thing your user would feel physically pained to lose? If you can't answer in five seconds, you're about to spend six months building features nobody asked for.
For the 47-row founder, the answer turned out to be a single button, "price my project." Forty-six other features were scaffolding for that one moment.
2. The Un-Interview
Most founder "user research" is a confession booth. They show a prototype, ask "would you use this?", and collect the polite yeses they came for. It's a waste of 20 calls.
We run un-interviews. The rule: you never mention your solution. You ask the user to describe the last time the problem made them want to throw their laptop out the window. You're not a salesperson, you're an investigator. They are symptom reporters, not solution designers. The Voice of the Customer is not an instruction manual. It's a map of pain.
If you can't write your user's headache in one sentence after twenty calls, you haven't listened long enough.
3. The Ethical Smoke Test
Demand is a hallucination until someone gives you money, time, or a real intent signal. Before we touch a repo, we put up a smoke test landing page, the product written as if it already exists, with a real promise and a real price.
When a user clicks Buy, they get the truth: "We're not live yet. Want to be first in?" No tricks. No bait. Just a clean signal.
The numbers tell you what to do next:
3–5% conversion — weak. Pivot or re-frame.
10–20% conversion — green light. Build the photo filter.
>20% conversion — warning bell. The pull is so strong that if your backend is fragile, it'll catch fire on day one.
That last one is the trap we see most often. Founders treat virality as a finish line. It's an ignition source.
4. The Rebuild-Ready Foundation
Most MVPs don't die from lack of features. They die the week traction finally arrives, when the architecture chosen for "just validating" collapses under the first real load. We've inherited too many of those repos to think it's bad luck.
So we tell every founder the same thing: ship fast, but architect like you'll be glad you did in month six. A real relational data model from day one. Clear boundaries between the product and the integrations bolted onto it. Automated tests for the one workflow that actually matters, the photo filter from cut #1. CI/CD before you have anyone to deploy for.
This isn't gold-plating, and it isn't premature optimisation. It's the difference between a launch that compounds and a launch that has to be rebuilt mid-fire. Rebuild-ready isn't pessimism. It's the only posture that survives the moment your smoke test was right.
5. The Disappointment Metric
Sign-ups are a vanity metric. So is GMV in week one. So is press. The only PMF question that matters is the Sean Ellis test:
How would you feel if you could no longer use this product?
If 40% or more of your active users say "very disappointed," you've hit Product-Market Fit. If they shrug, you haven't — no matter how good the chart looks. PMF isn't a switch. It's a spectrum, and it moves. A competitor ships a delighter, expectations reset, and your "fit" decays. That's why retention curves matter more than acquisition charts. Flat retention is the only honest signal.
Conclusion: The Only Button That Matters
The 47-feature founder shipped six weeks later. One screen. One button. Price my project. Forty-three of his original rows never got built — and never needed to. The product is live. He's paying himself. His wife stopped asking.
An MVP isn't a destination. It's an experiment whose only job is to earn you a second date with your user. Speed of execution is your only real competitive advantage. Perfectionism is just a delay tactic for founders afraid to be wrong in public.
So here's the question we leave every founder with, the same one we asked across that table, six weeks before the thing went live:
If you had to launch tomorrow with one button, what would that button do?
Go build that button. Done is the only thing that will ever teach you how to be perfect.
title:
Your MVP Is a Graveyard
date:
[topics]
Product Strategy
Founders
Product Market Fit
[blog]
//More Blogs
//
Your MVP Is a Graveyard
Most founders bury their product before it ever ships. Here's how we keep them above ground.
//Enterprise Software
[overview]
title:
Your MVP Is a Graveyard
date:
[topics]
Product Strategy
Founders
Product Market Fit
The 47-Feature MVP
The Notion doc was forty-seven rows long.
The founder slid it across the table, toggles, dashboards, a billing portal, a "social tab," dark mode, three onboarding flows, a referral engine, an admin panel nobody had defined. Every row had a checkbox. Every checkbox was unticked. He called it the "MVP scope."
He had six weeks of runway, one developer he hadn't hired yet, and a wife asking him when the thing was going to make money.
We didn't argue. We just turned the doc around and started crossing things out with a red pen. By the time we got to row twenty-three he stopped us. "Wait, if you cut all of that, what's even left?"
That's the question. That's always the question. And the answer is the only thing that ever mattered.
TL;DR, If your "MVP" has more than ten features, you're not building a product, you're digging a graveyard. The fix isn't more discipline, it's a smaller, sharper bet: one feature people would scream about losing, conversations that don't mention your solution, and a smoke test that costs less than your domain renewal. Speed is the strategy. Done is the teacher.
The Problem: Founders Confuse Completeness With Confidence
Every founder we meet has the same hallucination. A perfect launch day. A polished platform. Users marvelling at the settings menu. Press. A graph that goes up and to the right.
What they actually have is a backlog. A long, beautifully-organised, completely unvalidated backlog. The features feel like progress because they look like a product. But every row in that doc is an unverified assumption dressed up as a deliverable.
Here's the failure mode, named plainly: founders build to feel safe, not to learn fast. A 47-row MVP isn't ambition, it's avoidance. It's the founder buying themselves three more months of "we're not ready to launch yet" before the market has a chance to tell them they're wrong.
The cost isn't just time. It's worse. Every extra feature dilutes the one thing the product is supposed to be unmistakably good at. By the time you ship, your core value is buried under a mountain of nice-to-haves. The user doesn't know what you're for. Neither do you anymore.
The agencies don't help, they quote against the 47-row doc, because that's how they bill. The no-code influencers don't help, they'll happily sell you the stack to build all of it. And the founder community doesn't help, because everyone's quietly comparing scope and nobody wants to be the one who shipped "less."
The graveyard is full of products that were almost ready.
How We Do It: The Five Cuts We Make Before A Line of Code Ships
When that founder asked what was even left after we crossed out twenty-three rows, we showed him the actual MVP framework we use at the studio. Five cuts. In this order. No skipping.
1. The Photo Filter Test
Instagram didn't launch as a social network. It launched as a filter app. No comments. No stories. No reels. One job: make a phone photo look better. They added everything else after people would have walked into traffic to keep using it.
So we ask every founder: what's your photo filter? The one thing your user would feel physically pained to lose? If you can't answer in five seconds, you're about to spend six months building features nobody asked for.
For the 47-row founder, the answer turned out to be a single button, "price my project." Forty-six other features were scaffolding for that one moment.
2. The Un-Interview
Most founder "user research" is a confession booth. They show a prototype, ask "would you use this?", and collect the polite yeses they came for. It's a waste of 20 calls.
We run un-interviews. The rule: you never mention your solution. You ask the user to describe the last time the problem made them want to throw their laptop out the window. You're not a salesperson, you're an investigator. They are symptom reporters, not solution designers. The Voice of the Customer is not an instruction manual. It's a map of pain.
If you can't write your user's headache in one sentence after twenty calls, you haven't listened long enough.
3. The Ethical Smoke Test
Demand is a hallucination until someone gives you money, time, or a real intent signal. Before we touch a repo, we put up a smoke test landing page, the product written as if it already exists, with a real promise and a real price.
When a user clicks Buy, they get the truth: "We're not live yet. Want to be first in?" No tricks. No bait. Just a clean signal.
The numbers tell you what to do next:
3–5% conversion — weak. Pivot or re-frame.
10–20% conversion — green light. Build the photo filter.
>20% conversion — warning bell. The pull is so strong that if your backend is fragile, it'll catch fire on day one.
That last one is the trap we see most often. Founders treat virality as a finish line. It's an ignition source.
4. The Rebuild-Ready Foundation
Most MVPs don't die from lack of features. They die the week traction finally arrives, when the architecture chosen for "just validating" collapses under the first real load. We've inherited too many of those repos to think it's bad luck.
So we tell every founder the same thing: ship fast, but architect like you'll be glad you did in month six. A real relational data model from day one. Clear boundaries between the product and the integrations bolted onto it. Automated tests for the one workflow that actually matters, the photo filter from cut #1. CI/CD before you have anyone to deploy for.
This isn't gold-plating, and it isn't premature optimisation. It's the difference between a launch that compounds and a launch that has to be rebuilt mid-fire. Rebuild-ready isn't pessimism. It's the only posture that survives the moment your smoke test was right.
5. The Disappointment Metric
Sign-ups are a vanity metric. So is GMV in week one. So is press. The only PMF question that matters is the Sean Ellis test:
How would you feel if you could no longer use this product?
If 40% or more of your active users say "very disappointed," you've hit Product-Market Fit. If they shrug, you haven't — no matter how good the chart looks. PMF isn't a switch. It's a spectrum, and it moves. A competitor ships a delighter, expectations reset, and your "fit" decays. That's why retention curves matter more than acquisition charts. Flat retention is the only honest signal.
Conclusion: The Only Button That Matters
The 47-feature founder shipped six weeks later. One screen. One button. Price my project. Forty-three of his original rows never got built — and never needed to. The product is live. He's paying himself. His wife stopped asking.
An MVP isn't a destination. It's an experiment whose only job is to earn you a second date with your user. Speed of execution is your only real competitive advantage. Perfectionism is just a delay tactic for founders afraid to be wrong in public.
So here's the question we leave every founder with, the same one we asked across that table, six weeks before the thing went live:
If you had to launch tomorrow with one button, what would that button do?
Go build that button. Done is the only thing that will ever teach you how to be perfect.
//
Your MVP Is a Graveyard
Most founders bury their product before it ever ships. Here's how we keep them above ground.
//Enterprise Software
[overview]
The 47-Feature MVP
The Notion doc was forty-seven rows long.
The founder slid it across the table, toggles, dashboards, a billing portal, a "social tab," dark mode, three onboarding flows, a referral engine, an admin panel nobody had defined. Every row had a checkbox. Every checkbox was unticked. He called it the "MVP scope."
He had six weeks of runway, one developer he hadn't hired yet, and a wife asking him when the thing was going to make money.
We didn't argue. We just turned the doc around and started crossing things out with a red pen. By the time we got to row twenty-three he stopped us. "Wait, if you cut all of that, what's even left?"
That's the question. That's always the question. And the answer is the only thing that ever mattered.
TL;DR, If your "MVP" has more than ten features, you're not building a product, you're digging a graveyard. The fix isn't more discipline, it's a smaller, sharper bet: one feature people would scream about losing, conversations that don't mention your solution, and a smoke test that costs less than your domain renewal. Speed is the strategy. Done is the teacher.
The Problem: Founders Confuse Completeness With Confidence
Every founder we meet has the same hallucination. A perfect launch day. A polished platform. Users marvelling at the settings menu. Press. A graph that goes up and to the right.
What they actually have is a backlog. A long, beautifully-organised, completely unvalidated backlog. The features feel like progress because they look like a product. But every row in that doc is an unverified assumption dressed up as a deliverable.
Here's the failure mode, named plainly: founders build to feel safe, not to learn fast. A 47-row MVP isn't ambition, it's avoidance. It's the founder buying themselves three more months of "we're not ready to launch yet" before the market has a chance to tell them they're wrong.
The cost isn't just time. It's worse. Every extra feature dilutes the one thing the product is supposed to be unmistakably good at. By the time you ship, your core value is buried under a mountain of nice-to-haves. The user doesn't know what you're for. Neither do you anymore.
The agencies don't help, they quote against the 47-row doc, because that's how they bill. The no-code influencers don't help, they'll happily sell you the stack to build all of it. And the founder community doesn't help, because everyone's quietly comparing scope and nobody wants to be the one who shipped "less."
The graveyard is full of products that were almost ready.
How We Do It: The Five Cuts We Make Before A Line of Code Ships
When that founder asked what was even left after we crossed out twenty-three rows, we showed him the actual MVP framework we use at the studio. Five cuts. In this order. No skipping.
1. The Photo Filter Test
Instagram didn't launch as a social network. It launched as a filter app. No comments. No stories. No reels. One job: make a phone photo look better. They added everything else after people would have walked into traffic to keep using it.
So we ask every founder: what's your photo filter? The one thing your user would feel physically pained to lose? If you can't answer in five seconds, you're about to spend six months building features nobody asked for.
For the 47-row founder, the answer turned out to be a single button, "price my project." Forty-six other features were scaffolding for that one moment.
2. The Un-Interview
Most founder "user research" is a confession booth. They show a prototype, ask "would you use this?", and collect the polite yeses they came for. It's a waste of 20 calls.
We run un-interviews. The rule: you never mention your solution. You ask the user to describe the last time the problem made them want to throw their laptop out the window. You're not a salesperson, you're an investigator. They are symptom reporters, not solution designers. The Voice of the Customer is not an instruction manual. It's a map of pain.
If you can't write your user's headache in one sentence after twenty calls, you haven't listened long enough.
3. The Ethical Smoke Test
Demand is a hallucination until someone gives you money, time, or a real intent signal. Before we touch a repo, we put up a smoke test landing page, the product written as if it already exists, with a real promise and a real price.
When a user clicks Buy, they get the truth: "We're not live yet. Want to be first in?" No tricks. No bait. Just a clean signal.
The numbers tell you what to do next:
3–5% conversion — weak. Pivot or re-frame.
10–20% conversion — green light. Build the photo filter.
>20% conversion — warning bell. The pull is so strong that if your backend is fragile, it'll catch fire on day one.
That last one is the trap we see most often. Founders treat virality as a finish line. It's an ignition source.
4. The Rebuild-Ready Foundation
Most MVPs don't die from lack of features. They die the week traction finally arrives, when the architecture chosen for "just validating" collapses under the first real load. We've inherited too many of those repos to think it's bad luck.
So we tell every founder the same thing: ship fast, but architect like you'll be glad you did in month six. A real relational data model from day one. Clear boundaries between the product and the integrations bolted onto it. Automated tests for the one workflow that actually matters, the photo filter from cut #1. CI/CD before you have anyone to deploy for.
This isn't gold-plating, and it isn't premature optimisation. It's the difference between a launch that compounds and a launch that has to be rebuilt mid-fire. Rebuild-ready isn't pessimism. It's the only posture that survives the moment your smoke test was right.
5. The Disappointment Metric
Sign-ups are a vanity metric. So is GMV in week one. So is press. The only PMF question that matters is the Sean Ellis test:
How would you feel if you could no longer use this product?
If 40% or more of your active users say "very disappointed," you've hit Product-Market Fit. If they shrug, you haven't — no matter how good the chart looks. PMF isn't a switch. It's a spectrum, and it moves. A competitor ships a delighter, expectations reset, and your "fit" decays. That's why retention curves matter more than acquisition charts. Flat retention is the only honest signal.
Conclusion: The Only Button That Matters
The 47-feature founder shipped six weeks later. One screen. One button. Price my project. Forty-three of his original rows never got built — and never needed to. The product is live. He's paying himself. His wife stopped asking.
An MVP isn't a destination. It's an experiment whose only job is to earn you a second date with your user. Speed of execution is your only real competitive advantage. Perfectionism is just a delay tactic for founders afraid to be wrong in public.
So here's the question we leave every founder with, the same one we asked across that table, six weeks before the thing went live:
If you had to launch tomorrow with one button, what would that button do?
Go build that button. Done is the only thing that will ever teach you how to be perfect.
title:
Your MVP Is a Graveyard
date:
[topics]
Product Strategy
Founders
Product Market Fit
//
Your MVP Is a Graveyard
Most founders bury their product before it ever ships. Here's how we keep them above ground.
//Enterprise Software
[overview]
The 47-Feature MVP
The Notion doc was forty-seven rows long.
The founder slid it across the table, toggles, dashboards, a billing portal, a "social tab," dark mode, three onboarding flows, a referral engine, an admin panel nobody had defined. Every row had a checkbox. Every checkbox was unticked. He called it the "MVP scope."
He had six weeks of runway, one developer he hadn't hired yet, and a wife asking him when the thing was going to make money.
We didn't argue. We just turned the doc around and started crossing things out with a red pen. By the time we got to row twenty-three he stopped us. "Wait, if you cut all of that, what's even left?"
That's the question. That's always the question. And the answer is the only thing that ever mattered.
TL;DR, If your "MVP" has more than ten features, you're not building a product, you're digging a graveyard. The fix isn't more discipline, it's a smaller, sharper bet: one feature people would scream about losing, conversations that don't mention your solution, and a smoke test that costs less than your domain renewal. Speed is the strategy. Done is the teacher.
The Problem: Founders Confuse Completeness With Confidence
Every founder we meet has the same hallucination. A perfect launch day. A polished platform. Users marvelling at the settings menu. Press. A graph that goes up and to the right.
What they actually have is a backlog. A long, beautifully-organised, completely unvalidated backlog. The features feel like progress because they look like a product. But every row in that doc is an unverified assumption dressed up as a deliverable.
Here's the failure mode, named plainly: founders build to feel safe, not to learn fast. A 47-row MVP isn't ambition, it's avoidance. It's the founder buying themselves three more months of "we're not ready to launch yet" before the market has a chance to tell them they're wrong.
The cost isn't just time. It's worse. Every extra feature dilutes the one thing the product is supposed to be unmistakably good at. By the time you ship, your core value is buried under a mountain of nice-to-haves. The user doesn't know what you're for. Neither do you anymore.
The agencies don't help, they quote against the 47-row doc, because that's how they bill. The no-code influencers don't help, they'll happily sell you the stack to build all of it. And the founder community doesn't help, because everyone's quietly comparing scope and nobody wants to be the one who shipped "less."
The graveyard is full of products that were almost ready.
How We Do It: The Five Cuts We Make Before A Line of Code Ships
When that founder asked what was even left after we crossed out twenty-three rows, we showed him the actual MVP framework we use at the studio. Five cuts. In this order. No skipping.
1. The Photo Filter Test
Instagram didn't launch as a social network. It launched as a filter app. No comments. No stories. No reels. One job: make a phone photo look better. They added everything else after people would have walked into traffic to keep using it.
So we ask every founder: what's your photo filter? The one thing your user would feel physically pained to lose? If you can't answer in five seconds, you're about to spend six months building features nobody asked for.
For the 47-row founder, the answer turned out to be a single button, "price my project." Forty-six other features were scaffolding for that one moment.
2. The Un-Interview
Most founder "user research" is a confession booth. They show a prototype, ask "would you use this?", and collect the polite yeses they came for. It's a waste of 20 calls.
We run un-interviews. The rule: you never mention your solution. You ask the user to describe the last time the problem made them want to throw their laptop out the window. You're not a salesperson, you're an investigator. They are symptom reporters, not solution designers. The Voice of the Customer is not an instruction manual. It's a map of pain.
If you can't write your user's headache in one sentence after twenty calls, you haven't listened long enough.
3. The Ethical Smoke Test
Demand is a hallucination until someone gives you money, time, or a real intent signal. Before we touch a repo, we put up a smoke test landing page, the product written as if it already exists, with a real promise and a real price.
When a user clicks Buy, they get the truth: "We're not live yet. Want to be first in?" No tricks. No bait. Just a clean signal.
The numbers tell you what to do next:
3–5% conversion — weak. Pivot or re-frame.
10–20% conversion — green light. Build the photo filter.
>20% conversion — warning bell. The pull is so strong that if your backend is fragile, it'll catch fire on day one.
That last one is the trap we see most often. Founders treat virality as a finish line. It's an ignition source.
4. The Rebuild-Ready Foundation
Most MVPs don't die from lack of features. They die the week traction finally arrives, when the architecture chosen for "just validating" collapses under the first real load. We've inherited too many of those repos to think it's bad luck.
So we tell every founder the same thing: ship fast, but architect like you'll be glad you did in month six. A real relational data model from day one. Clear boundaries between the product and the integrations bolted onto it. Automated tests for the one workflow that actually matters, the photo filter from cut #1. CI/CD before you have anyone to deploy for.
This isn't gold-plating, and it isn't premature optimisation. It's the difference between a launch that compounds and a launch that has to be rebuilt mid-fire. Rebuild-ready isn't pessimism. It's the only posture that survives the moment your smoke test was right.
5. The Disappointment Metric
Sign-ups are a vanity metric. So is GMV in week one. So is press. The only PMF question that matters is the Sean Ellis test:
How would you feel if you could no longer use this product?
If 40% or more of your active users say "very disappointed," you've hit Product-Market Fit. If they shrug, you haven't — no matter how good the chart looks. PMF isn't a switch. It's a spectrum, and it moves. A competitor ships a delighter, expectations reset, and your "fit" decays. That's why retention curves matter more than acquisition charts. Flat retention is the only honest signal.
Conclusion: The Only Button That Matters
The 47-feature founder shipped six weeks later. One screen. One button. Price my project. Forty-three of his original rows never got built — and never needed to. The product is live. He's paying himself. His wife stopped asking.
An MVP isn't a destination. It's an experiment whose only job is to earn you a second date with your user. Speed of execution is your only real competitive advantage. Perfectionism is just a delay tactic for founders afraid to be wrong in public.
So here's the question we leave every founder with, the same one we asked across that table, six weeks before the thing went live:
If you had to launch tomorrow with one button, what would that button do?
Go build that button. Done is the only thing that will ever teach you how to be perfect.
title:
Your MVP Is a Graveyard
date:
[topics]
Product Strategy
Founders
Product Market Fit
[blog]