Founder Reads
Nine things I'd hand anyone building a company alone. Six essays, a guide, a talk and a book, each with the one idea worth taking from it.
One email unlocks the whole list, and every other guide on the site. Every link goes to the original, which is free to read.
All Nine, At A Glance
Roughly in the order I'd read them. Click a row for my notes, or go straight to the original.
Do Things That Don't Scale
▶ Paul Graham · Jul 2013
Startups don't take off on their own. Graham's case is that you go and get your first users, one at a time, and then look after each of them so well they tell other people. Recruiting by hand, over-serving early users and starting with a market small enough to win are all things that won't scale. That's the point.
Users after a year at 10% a week, starting from 100.
Users after two years at the same rate.
Pebble's Kickstarter, after the founders hand-assembled the first watches.
- Recruit by hand. Airbnb's founders went door to door in New York. Stripe set new users up on the spot, on their own laptops.
- Delight, don't just serve. Your first users should feel that signing up was one of the best choices they ever made.
- Start narrow. Facebook began with one college. A small market you can saturate beats a big one you barely touch.
- Do the work yourself. The manual version teaches you what's worth automating later.
Personally recruit five users, and fix the first thing each of them complains about.
“Actually startups take off because the founders make them take off.”Paul Graham
- Author
- Paul Graham
- Published
- Jul 2013
- Format
- Essay
- Where
- paulgraham.com
The User Manual
▶ Amy Buechler · Scale Yourself
Amy Buechler is a licensed psychotherapist and was YC's first in-house founder coach. The User Manual is her guide to the inside of a founder's head: the pressure, anxiety and self-doubt that come with the job, and practical ways to work with them instead of being run by them. When you build alone there's no cofounder to notice you're off, so read this before you need it.
- Reactivity. Catching yourself reacting when you meant to choose.
- Burnout. Seeing it early, before it starts making your decisions.
- Fear of losing momentum. The pressure to keep moving, and what it costs.
- Building trust. With yourself first, then with the people around you.
Start with the chapter that made you wince when you read that list.
“You are the most important creation of your startup.”Amy Buechler
- Author
- Amy Buechler
- Format
- Online Guide
- Where
- scaleyourself.com
Product-Market Fit Isn't Enough
▶ Brian Balfour's Four Fits · by Garrett Jestice · Apr 2025
Garrett Jestice breaks down Brian Balfour's Four Fits framework. Product-market fit is only the first of four fits that have to line up before growth compounds: the market has to fit the product, the product has to fit a channel, the channel has to fit your business model, and the model has to fit the market. Miss one and growth stalls, however much people like what you built.
- Market–Product. A defined group of people has the problem, and your product solves it.
- Product–Channel. The product suits the channel people will find it through.
- Channel–Model. What a customer costs to win in that channel works with what you charge.
- Model–Market. Your pricing and model suit how big the market is and how it buys.
Write one sentence for each fit. The one you can't write is your bottleneck.
“Product-market fit is just the first checkpoint.”Garrett Jestice
- Author
- Garrett Jestice
- Framework
- Brian Balfour
- Published
- Apr 8, 2025
- Where
- GTM Foundations
How To Keep Your Users
▶ David Lieb · YC Startup School
YC Group Partner David Lieb on cohort retention: the share of new users who keep coming back. It's the clearest early answer to YC's question of whether you made something people want. He covers how to define cohorts, which action counts as active, what a good curve looks like, the ways founders fool themselves, and how to improve the curve once you can see it.
- Cohorts, not totals. Group users by the week they joined instead of watching one blended number.
- Pick a real action. Count the thing that means value, like a completed ride, not a login.
- Look for the flattening. A curve that levels off means people stay. One that slides to zero means you're not there yet.
- Then improve it. Smoother onboarding, a faster path to the "aha", and calling the users who left.
Chart your weekly cohorts, using one action that means real value.
“At YC our motto is make something people want.”From The Video's Description
- Speaker
- David Lieb
- Series
- YC Startup School
- Length
- About 30 Min
- Where
- YouTube
The 4 Types Of Defensibility
▶ James Currier · NFX
Why Amazon and Yahoo couldn't take eBay down, even with free listings and better technology. James Currier walks through the four defenses a tech company can have, and why brand and scale have weakened while network effects stay close to unbeatable.
- Brand. Recognition and loyalty. Real, but Lotus 1-2-3 had it and still lost to Microsoft.
- Scale. Fixed costs spread over more customers. Weaker now that big tech can outspend anyone.
- Embedding. Being wired into how customers work, so switching hurts.
- Network effects. Each new user makes it better for the rest. It's how eBay held off free clones.
Name your moat in one word. If you can't, that's the next thing to build.
“If you don't have a durable advantage, well, you're roadkill.”From The Essay
- Author
- James Currier
- Format
- Essay
- Where
- nfx.com
Default Alive Or Default Dead?
▶ Paul Graham · Oct 2015
One question: on your current expenses and growth rate, do you reach profitability on the money you have left? Graham says half the founders he talks to don't know. If the answer is no, you're default dead, and the time to change course is now, not after the next raise falls through.
- Know which one you are. Hold expenses flat, keep revenue growing at its recent rate, and see whether you get there.
- The fatal pinch. Default dead, plus slow growth, plus not enough time left to fix it.
- Don't plan on the next raise. Treat new money as a bonus, and keep a plan that works without it.
- Hire slowly. Growth early on comes from the product, not the headcount.
Write down four numbers: cash, monthly burn, monthly revenue and its growth rate. Then do the math.
“Hiring too fast is by far the biggest killer of startups that raise money.”Paul Graham
- Author
- Paul Graham
- Published
- Oct 2015
- Format
- Essay
- Where
- paulgraham.com
Startup = Growth
▶ Paul Graham · Sep 2012
What makes a startup a startup isn't being new or being tech. It's being built to grow fast. Graham argues the growth rate is your compass: pick the number that matters, set a weekly target, and let hitting it decide what you work on next.
A year of 1% weekly growth.
A year at 5% a week.
A year at 10% a week.
- Built to grow. A startup is designed to grow fast, not just young or technical.
- Two problems. Make something lots of people want, then reach all of them.
- Know your number. Without it you can't tell whether you're doing well or badly.
- Small rates compound. 1% versus 5% a week is the gap between 1.7x and 12.6x a year.
Pick one weekly number (revenue if you charge, active users if you don't) and write it down every Monday.
“A startup is a company designed to grow fast.”Paul Graham
- Author
- Paul Graham
- Published
- Sep 2012
- Format
- Essay
- Where
- paulgraham.com
The Product Death Cycle
▶ David J. Bland · Discussed By Dan Olsen
A loop most products fall into at least once. Nobody uses the product, so the team asks customers what's missing, builds those features, and nobody uses those either. Repeat until the money runs out. David J. Bland drew the cycle, and product author Dan Olsen is among those who use it to explain why taking feature orders isn't the same as understanding the problem.
- Problems, not solutions. Customers describe pain well and design fixes badly. Dig for the pain.
- Test before you build. Prototypes and MVPs check value and usability before you write code.
- Work from a strategy. Tie each feature to an outcome you're trying to move, not a request list.
- Push back. Decide on what customers do, not on the loudest opinion.
Before you build the next feature request, ask what the person was trying to do when they hit the wall.
No one uses it. Ask what's missing. Build it. Repeat.The Cycle, Paraphrased
- Coined By
- David J. Bland
- Discussed By
- Dan Olsen
- Format
- Concept
The Mom Test
▶ Rob Fitzpatrick · 2013
The subtitle is the whole book: how to talk to customers and learn if your business is a good idea when everyone is lying to you. People are polite about your idea, especially people who like you. Fitzpatrick's rules make it hard for them to be polite, because you stop asking about your idea and start asking about their life.
- Their life, not your idea. Talk about their life instead of your idea.
- The past, not the future. Ask about specifics in the past instead of generics or opinions about the future.
- Listen. Talk less and listen more.
- Compliments aren't data. "I'd use that" costs nothing to say. What people already do, and pay for, is the evidence.
Rewrite your interview questions so none of them mention your product.
“Talk about their life instead of your idea.”Rob Fitzpatrick
- Author
- Rob Fitzpatrick
- Published
- 2013
- Format
- Book
- Where
- momtestbook.com
I'm Building A Community For People Building Alone.
It isn't open yet. The waitlist is.
Much of this list comes down to one point: the code is the easy part now, and the hard parts are easier with other people around. I'm putting together a place for solo founders and vibe coders to compare notes on what's working, get honest feedback on what they're building, and find their first customers without guessing. Join the waitlist and you'll hear first when the doors open.
You're On The List ✓
I'll write once, when it's ready. Nothing else.