Ownership

Build the tool you actually need

An app I built for myself failed me at a festival, and one I built for myself held up on stage. The difference is what you owe people once you sell it.

5 min read8

Last week I was at LISTEN, a music festival, sitting with my laptop open in the afternoon, trying to put together a USB stick for a set I was playing later that night. I had produced a new song and I really wanted to play it. I didn't get to.

The app I was exporting from is called Crate, and I built it. It's my own audio toolbox. I've been collecting music my whole life, so I have a lot of audio, everything from a single hi-hat sample to an hour-long sound bath, and Crate is where all of it lives. It also has a simple field recorder and a virtual patch bay, so I can route audio from one app to another. One of the main reasons I built it was to manage my library and export DJ USB sticks from the same place, without having to submit to rekordbox, the Pioneer software that most club gear expects.

And I had never once tested that export. I built Crate about a month ago, I didn't have a gig coming up, so I just never did it. When I finally tried it at the festival, it had made three copies of everything in rekordbox, because I hadn't opened rekordbox in a month. Now it's a mess, and the new song stayed on my laptop.

Necessity

I make so much shit that sometimes it doesn't get tested, and that's something I'm really struggling with right now. I think it's the problem with a lot of this vibe coding. The stuff we make has to be used, and it's best if it's used by us. The best person you can make a tool for is yourself, so if you're making a tool, make one that you actually need.

If you look back at human history, the most important technology came out of necessity. Somebody faced a real problem at a certain time and solved it. Crate came out of a real need too. I just didn't use it for the thing it was for until the moment it mattered, and part of that is a tension I have as a creative: I don't always give myself the time to finish a project, because there's another project, or something else comes up.

If it had been somebody else's library, it would have been a nightmare, and they'd have lost all trust in the app. A lot of people don't trust rekordbox, but at least it's functional. Crate wasn't, at the one moment that counted. That was definitely a lesson learned.

The one I'd take on stage

Compare that to Cloister. Cloister is an effects rack for the iPad, and I built it so I could use it in a specific live performance. That's so clear to me. It hits everything: I enjoy making the thing, I enjoy performing, I enjoy music, and I enjoy technology.

I've been making things with AI all year, and Cloister is the only one I've been stoked enough on to really think about releasing. I trusted it to get up on stage in front of a whole bunch of people and get after it. Sure, it's vibe coded, and it might not be the most stable thing on planet Earth. I don't fully know how it works. What I do know is that it worked in that moment, and that's better than user feedback for me, because I know it works in my workflow.

That's what I would call an aligned offer. It's something I can stand behind, something I use, something I even made, and something I know fits a very specific moment.

What changes when someone else uses it

There's a huge difference between a personal environment and a production environment. When it's mine, I know what it takes to set it up. Honestly, the first 15 to 20 minutes of using any of my tools is fixing them or improving them, and that's usually how I get into the work. That's also how they get better, because my workflow changes, and everything is changing so fast that the tool ends up tailored to what I need at that exact moment rather than being a static product.

A stranger doesn't get any of that. How they use it is going to be different from how I use it, which leads to support requests, which leads to features, and then it's a product that's drifting away from what I built. Especially if somebody paid for it, and especially if it's a subscription.

I think about it like renting versus owning. If I rent an apartment, I expect my landlord to do basic maintenance. If I own the house, it's all on me. A subscription is rent: somebody keeps paying, so they expect it to be maintained, improved, and fixed when something breaks. Every object we use has some obligation to maintain it if we want it to keep working, like a car, and I don't think people realize that about software, especially the software they make for other people. There are so many vibe-coded tools out there that aren't maintained, aren't functional, and were made for a quick buck because somebody thought it was cool. I do feel like there's a responsibility here.

Maybe there's a version where the code lives open source on GitHub and there's a simpler version in the app store, and you're paying for the wrapper, the easy download and the distribution. Maybe that's a way to do it: you pay for it once, or you pay per build.

Say how you made it

Transparency in how things are made, and what the limits are, is really important to me, and I'm not feeling it from the world right now. Everything feels like snake oil. Everyone's got a vibe-coded tool, and there's no acknowledgement of "this is how I made this, this is where I'm at in the process, this is what it takes to keep working on it, and that's why it's $5." You could spend that money on tokens and build it yourself, or you could just buy the app.

I also have to acknowledge that the tools I make are built on tools that already exist. I didn't come up with those ideas out of the blue. I put them together in new ways. Everything's a remix, and there's value in packaging things up and being open about that. At that point, sure, you can sell it, but you've got to be honest about what you're selling and what the future is for it.

What people are paying for is the ease and the access to something that helps them get to whatever they're working towards, and the understanding that even if it's not perfect, it's going to be okay. I've spent a lot of money over the years on stuff I don't use anymore. Musicians call that gear acquisition syndrome, buying the thing instead of making something helpful, and maybe some of what I build is just more of it. There has to be a moment the tool is meant to serve. It has to be a win-win, where it helps someone earn a little or save some time, rather than me telling you "this is going to make your life better." I can make you feel good for 15 minutes, but it's just not going to work.

Who this is for

This is for anyone refining their process, whether that's a day-to-day workflow or a very specific way of making something. You have to understand how each tool works in your own system, kind of like an assembly line. Or a drum set: there are a whole bunch of drums available, but I only have two arms and two legs, so I can only hit four of them at once. You have to know what your workflow is, and what you're going to use the thing for.

Building custom tools is a great way to figure out what you're actually trying to make. We live in a time where anybody can create pretty much anything, and that's only speeding up. There's no better way to understand how you work than to create the tool that you use to work, and it's a positive feedback loop.

David Kerns

David Kerns

I help people bring their visions to life. Sharing thoughts on operations, product, and making things that matter.

More about David

Enjoyed this post?

Subscribe to get new posts delivered to your inbox. No spam.