The best companies don't compete head-on

I'm always curious how great companies get built. Not the funding rounds or founder mythology, but the product decisions.

The pattern that keeps showing up is simple: the best companies usually do not win by competing head-on. They change the category.

Peter Thiel writes about this in Zero to One. His argument is that going from 0 to 1, creating something new, is a completely different game from going from 1 to n, copying something that already works.

Competition feels safer because the market is already there. You can point to demand. You can compare features. You can tell yourself the plan is to be 10% better.

The problem is that everyone else can tell themselves the same thing.

Stripe is a great example. Before Stripe, online payments were technically possible. PayPal existed. Authorize.Net existed. Banks had APIs, even if they were painful to use. If you had asked developers in 2010 whether online payments were "solved," a lot of them probably would have said yes.

The Collisons saw it differently. Accepting payments was possible, but the experience was miserable. Stripe did not feel like a slightly better PayPal. It felt like someone had removed a layer of pain developers had accepted as normal.

That is the 0 to 1 move.

Not: "how do we improve the existing thing?"

More like: "what would this look like if we started from scratch?"

That question is powerful because it removes the safety of comparison. If you are copying, at least you know what to copy. If you are rethinking something, you have to make a judgment call before the market agrees with you.

But that is also where the upside is. Crowded markets feel safe until you realize they are crowded for a reason. Everyone sees the same opportunity. Everyone copies the same playbook. Margins disappear.

What's something in your space that everyone accepts as "just how it works," but could probably be rethought from the ground up?
Hit reply and tell me. I read every response.

Have a great week,

Graham

This week is sponsored by Nimbalyst! If you've played with any of the AI tools, you know how many there are, how it's tough to switch between them and keep all the context of a project in one place. That's the problem that Nimbalyst is aiming to solve.

📚 Book Notes

Zero to One - Peter Thiel

I first read this in 2019 when I was working in product management and thinking about starting something.

The question that stuck with me was: "What important truth do very few people agree with you on?"

I did not have a good answer then. I'm not sure I have a perfect one now. But I do think the question is useful because it forces you out of lazy comparison. If your idea only makes sense because someone else already proved it, you may be closer to 1 to n than 0 to 1.

The line I still think about: every moment in business happens only once. The next Zuckerberg will not build a social network. The next Stripe will not look like Stripe.

📖 Featured Article

I am building a cloud

David Crawshaw is building his own cloud infrastructure from scratch. Not because AWS is useless. Because he thinks the assumptions underneath modern cloud computing are worth questioning.

That is the interesting part. The project is not "how do I compete with AWS?" It is closer to "what would cloud computing look like if we rebuilt it around different constraints?"

You do not have to agree with the whole bet to admire the posture.

From the Blog

📝 How Stripe Turned 7 Lines of Code Into $107 Billion
The obvious companion to this issue. Payments were not broken in the way people usually mean. They were just painful enough that everyone had stopped noticing.

📝 The Skills That Actually Matter in the Age of AI
When everyone has access to the same tools, the edge moves upstream. Taste, judgment, and deciding what is worth building matter more than another prompt trick.

🔗 Things I Found Interesting

Inside the company building America's first mail-order servant robot
1X Technologies is trying to put humanoid robots into homes. Whether or not this specific company wins, someone is going to create a new category here. The question is whether the first useful home robot looks like what people expect.

The 48 Laws of Twitter
Some of these are contrarian. Some are obvious. I disagreed with plenty of them and still bookmarked the whole thing, which is usually a good sign.

Over-editing refers to a model modifying code beyond what is necessary
A useful note on AI coding tools that change too much code. The best assistant is not always the one that does the most. Restraint is a feature.

Casa Balma Murada
Mesura's work is always worth looking at. This one creates something new from old materials instead of demolishing and starting over. A good architectural version of the same question: what is worth preserving, and what needs to be rethought?

📣 Quote

"The best entrepreneurs I've funded have always had a specific insight about a specific problem."
Peter Thiel