Scan time: 2-3 minutes / Read time: 3-5 minutes
⠀
Howdy rebel ideapreneurs 🦸♂️🦸♀️
⠀
🙋♂️ Ethan’s first issue that never goes out
Ethan is a product manager at a software company.
On the side, he’s starting a newsletter for first-time product managers.
His first issue has been finished for three weeks.
Every night he opens it, reads it again, and closes it.
He keeps picturing what one bad issue would cost him:
a typo his old boss spots, three people who unsubscribe, a coworker who thinks he’s showing off.
🧩 His struggle:
He treats every possible mistake as a loss he can never win back, so he never sends anything.
At lunch one day, a coworker shares a video of a talk in the team chat.
It’s the true story of a NASA engineer whose projects failed again and again,
and who still grew a YouTube channel to more than 57 million subscribers.
His best lesson came from a puzzle about a little car in a maze.
⠀
🎮 The NASA engineer who took away the points
Mark Rober joined NASA’s Jet Propulsion Laboratory in 2004.
He worked there as an engineer for nine years.
Seven of those years were on the team behind Curiosity, the rover NASA sent to Mars.
On the side, he started a YouTube channel.
He used real engineering to build big, fun things, like the world’s largest Super Soaker water gun.
Most builds took him two or three months.
One, a dartboard that tracks a dart as it flies, took him three years.
When his channel had about three million subscribers, he asked them to try a puzzle he’d built with a friend.
On the screen was a little car in a maze.
You moved it by snapping blocks of computer code together, then pressing Run.
He said he wanted to show that anyone, from any background, could learn to code.
About fifty thousand people tried it, alone at home, and nobody paid them to.
What Mark didn’t tell them was that there were two versions of the puzzle.
Everyone started with 200 points.
In one version, when the code failed, the screen said: “That didn’t work. Please try again.”
In the other, it said: “That didn’t work. You lost 5 points. You now have 195 points. Please try again.”
The points weren’t worth anything.
They were just a number on a screen.
Turns out the people who never lost points made almost two and a half times as many tries.
And more of them solved it: 68 out of every 100, against 52 out of every 100 for the people who lost points.
Mark named what he saw after a game he loved as a kid, Super Mario Bros.
In the game, you run and jump through levels to rescue a princess.
When you fall into a pit, you don’t feel stupid, and you don’t quit.
You think: next time, jump a little later, or with more speed.
His friends at school asked which level he’d reached, never how many times he’d fallen.
Mark called it the Super Mario Effect: keep your eyes on the princess, not on the pits.
He runs his own builds the same way.
When one fails, his question is: “OK, that sucked, but what did we learn from that?”
Today more than 57 million people subscribe to his YouTube channel.
He also runs CrunchLabs, a company that sells science kits by subscription.
Get this — in one year, he made thirty million dollars.
Mark stopped counting each failed build as points lost, and asked what it taught him instead.
⠀
🔥 The recipe
➡️ Take Away the Points
Mark took the points away from half of the fifty thousand players, and more of that half solved his puzzle.
This approach is called Take Away the Points.
You name the points you think a mistake will cost you.
Then you rewrite each one as “That didn’t work. Try again.”
In Mark’s puzzle, the points were the 5 a player lost with every failed try.
Take them away, and the same failed try just said: “That didn’t work. Please try again.”
For us, the points are things like a typo someone spots, or a few people who unsubscribe.
Mark didn’t ask his players to make fewer mistakes.
He made each mistake cost nothing, so they could afford to make more of them.
So when our first issue sits finished and unsent, it’s usually not the writing that’s stopping us.
It’s the points we think we’ll lose if it flops.
Take those away, and each week, count how many issues you sent, not how many did well.
⠀
🛠️ Ethan’s turn
Now, back to Ethan, who has kept his finished first issue on his laptop for three weeks.
That evening, he runs the prompt at the bottom of this email,
and tells it what he’s afraid one bad issue will cost him.
The reply from his AI sidekick is short:
Nobody is taking points away from you, so send it this week,
then send one every week after, and let each issue fix whatever the last one got wrong.
Ethan reads his issue one last time, and he presses Send.
⠀
🌅 Ten weeks later
Every Tuesday, Ethan sends a new issue, and there have been ten so far.
Issue three got zero replies, so he made issue four shorter.
Issue four got five replies.
About a hundred and forty people read his newsletter today.
His old boss subscribed.
Nobody has mentioned a typo.
Your first issue doesn’t need to be safe from every mistake.
A mistake only tells you what to try next.
Keep your eyes on the princess, not on the pits.
That’s it, my fellow outliers!
Yours ‘showing you that building a business is fun, not a drain’ Vijay Peduru 🦸♂️
⠀
⠀
⠀
⠀
⠀
🏄♀️ The prompt
The prompt below finds the points you’re afraid to lose, takes each one away, and gives you one thing to send this week.
🧰 What you’ll need
The thing you haven’t sent yet
- Paste it, upload it, or describe it in a line or two.
- e.g. My first newsletter issue, on how to run your first planning meeting as a new product manager. It has been finished for three weeks.What you’re afraid a flop will cost you
- A minute’s worth. Type it, or say it out loud with a voice tool, messy is fine.
- e.g. My old boss spots a typo. A few people unsubscribe. A coworker thinks I’m showing off.
🚀 How to run it
Open your AI sidekick, ChatGPT, Claude, or whichever one you use. The best is one that can already see your notes, wherever you keep them (on your computer, on the web, or in the cloud).
Swap the two examples, the text inside the curly brackets, for your own, then run it. If your AI sidekick can already see your notes, you can run it as is.
🧪 Want to test it first? Run it without changing anything, and it will use the examples.
CONTEXT
- Use my own material first. If you can see my notes and my unsent work, wherever they are stored, read them there. If I uploaded a file, read that. Only then use the examples below.
- Write in plain, friendly words, like a friend who wants me to ship.
- If you know who my readers are, keep them in mind. Otherwise assume first-time product managers.
- If I have replaced the examples, use only mine. If an example is still there, fill it in from what you already know about me, and mark it as your assumption. If you don't know anything about me yet, use the example exactly as written. If I dictated input 2, it will be messy; that is fine, do not ask me to clean it up.
- The one rule for this build: a miss costs nothing unless it does real, lasting harm. Be honest. If one of my fears is a real risk (private information, a promise I can't keep), say so and tell me how to fix it before I send. Everything else is a point that isn't real.
INPUTS
1. The thing I haven't sent yet (pasted, uploaded, or described):
{e.g. My first newsletter issue, on how to run your first planning meeting as a new product manager. It has been finished for three weeks.}
2. What I'm afraid a flop will cost me (typed or dictated, a minute's worth):
{e.g. My old boss spots a typo. A few people unsubscribe. A coworker thinks I'm showing off.}
END OF INPUTS
OUTPUT
1. My points: list every cost I named in input 2, one per line, plus any you can see in input 1 that I didn't say out loud.
2. For each point, say whether it is a real risk or a point that isn't real, in one sentence each.
3. Fix the real risks in input 1, if there are any, and show me only the lines you changed.
4. Rewrite each point that isn't real in this exact shape:
"That didn't work. Try again: [what I would change in the next one]."
5. Give me a weekly number of tries, not wins: how many things I should send each week, based on how much time input 1 suggests I have.
6. Finish with one line on its own: what to send today, and when.