Hacker News new | ask | show | jobs
by farceSpherule 329 days ago
Everyone is trying to "copy" Pieter Levels "success" which as of now is "brand effect."

The guy started his thing over a decade ago and people look at it now and think they can replicate it.

The stuff the guy codes is garbage and what he does is far from solving any problems.

And, I do not believe his revenue numbers. At all. But people on the Internet see some shit posted, believe it, and then compare themselves to it.

Gleaning anything from his "1 in a million success" is falling prey to survivorship bias.

4 comments

I agree with half your comment while disagreeing with the other half. Yes, it is very true that he now has a brand and can get users much more easily, and that trying to replicate his success is very survivorship bias heavy. However, if he hadn't been solving problems for people, he wouldn't have made the money he has in the first place, because no one would be buying his products (I, for example, bought NomadList a long time ago and met many people from it due to their Slack channels). And see my other comment about "garbage" code, it does not matter if they're making money, Levels is by his own proclamation not a software engineer, he uses code as a tool.
How are you defining garbage?
Probably because Levels says he codes each product in a single PHP file. But then again, there's a reason he's successful and the parent is not (at least to that same level, pun intended). Technologists think code is an end unto itself while true entrepreneurs that it's just a means to an end, and that the end itself is money (otherwise, why are you running a business? If you don't make money, it's simply a hobby).
>it's just a means to an end, and that the end itself is money

The end is also to create something useful for your customers. Hopefully.

single file 30,000 of PHP code.
That truly doesn't matter though. It's certainly not what I would do, but assuming the numbers he claims are genuine, it's hard evidence that your customers don't care how you wrote the software.
The feature of software being maintainable and reusable by people other than its original author is for the users. It can be easy to forget until it’s too late.
The users don't care about this.

I write my software with Haskell, NixOS, comprehensive testing, static analysis, linting, a great deal of care, etc.

If I had written everything in one big PHP file however, the difference to a paying customer would be exactly nil.

1 file can be very maintainable. Reuse is promoted by a well designed library and interfaces, which can also be in 1 file.

Only thing 1 file can't do is allow concurrent work if bad version control is used. If patches are tight with good patch theory, 1 file is fine again.

Modern programming languages have namespaces, catch type bugs, all of it works quite well with 1 file. Recent OAuth Cloudflare 1 file library (coded a bit with AI) is a breath of fresh air, 2600 lines of data types and behavior in one place.

like his flying simulator.
I think Pieter is actually legit, even though his bad takes on things started to increase in the last few years. But his products?

- Nomadlist solved a problem, especially back then: a social network for digital nomads. Nothing wrong with it.

- Remoteok is a job board, but niched down, which is completely ok

- photoai / interiorai: it pains me that an AI slop generator is obviously a viable business model, but apparently many people are willing to pay for this

The rest is indeed brand effect. That shitty flight simulator wouldn't have gotten any traction if it wasn't for being part of "a community".

But, and this is the most important thing I like about Pieter: he doesn't sell shovels in a gold rush. His products solve problems for quite different groups of people and X isn't necessarily the primary marketing platform for his stuff.

There are others in the indie hacker scene that are way more shady, because they make their money from products that sell that lifestyle first and foremost.

same. his websites are indeed garbage.