You don't need a big app development company to build something real. I know because I built Carrd, a one-page site builder, mostly on my own, and the thing that saved me wasn't a bigger team or a fancier stack. It was saying no to almost everything.
Here's how that played out, and what it might mean for you if you're weighing whether to hire an app development company or just start small yourself.
Why did I break the project into five pieces?
By my old standards, Carrd was huge. Big enough to be intimidating. So instead of staring at the whole mountain, I split it into five subprojects, the Generator, the Builder, the Dashboard, the Site, and the Paid Plan, and worked them one at a time.
That's the whole trick. A big project you can't start is worth less than a small one you can finish. Cutting it into five let me make progress without the entire thing looming over me every morning.
How do you pick something small enough to build alone?
I'd spent years designing and coding site templates, then building sites to hand them out. Good work, but it went routine. I wanted something new. The open question was what.
The answer came from staying close to what I already knew. Instead of chasing a wild, unfamiliar idea, I looked one step sideways from my existing skills, and the next logical project showed up. A builder for one-page profile sites. For a solo build, that niche checked the boxes. Simple, wide appeal, already proven popular. That made it the strongest bet I could place.
Any app development company will tell you scope creep is what kills projects. The fix isn't more people. It's picking a lane narrow enough that one person can actually see the end of it.
What does a narrow focus actually buy you?
The whole product is built around one pattern, a simple "stack" of elements. That's the site spec. You add elements, you arrange them, you customize the details that matter. Because everything lives inside that one idea, there's a natural limit to its complexity, for me building it, and for you using it.
That narrow focus is the hero here. It's why the thing stays easy to use instead of turning into another do-everything tool nobody can figure out.
- Simply built. One core pattern, not fifty features fighting each other.
- Customizable. Deep control on the details, without the overwhelm.
- Responsive. A mobile-friendly builder, not just mobile-friendly sites.
Did going framework-less pay off?
For the Builder, I skipped the popular frameworks and hand-rolled the frontend with plain JavaScript and a bit of jQuery. Risky? Sure. But it meant the tooling bent to fit the Builder instead of me bending the Builder to fit the tooling.
The bonus. I knew exactly how every part worked, which made bugs easier to chase and features easier to bolt on. It also forced me to learn things a framework would've quietly handled for me. Worth trying at least once, even if you never do it that way again.
The design came from a sketch made years earlier
I had a rough picture of how the Builder should look from back when I first dreamed up the project in early 2015. I did a mockup before I'd even started on the Generator. A few months of nudging it around, and I landed on something that looked a lot like the final UI.
The Dashboard followed the same instinct. Make the user's own sites the centerpiece. Big, auto-generated screenshots, front and center. The same screenshot-first idea that worked for the Builder worked again for the hub. Reuse what already works.
Even the name was a constraint problem
The name shows up inside people's own subdomain URLs, so it had to be short, memorable, and low-key enough not to stand out. Add the fact that snappy two-letter domain extensions are hard to grab, and naming turned into its own little puzzle. Carrd, at carrd.co, was the answer that fit.
What I'd tell anyone building solo — or hiring it out
If you take one thing from all this, take the list I wish someone had handed me:
- Don't try to do too much. Limiting the focus is the only reason one person could build this at all.
- Don't obsess over competitors. Watch them too closely and you'll box yourself into their shape.
- Stick to your vision, but stay a little flexible. A user asking for multiple pages is what led me to the Control Element — a good idea I hadn't planned for.
- Don't force a signup just to try the thing. Let people play first.
- Plan away from the screen. Paper and a whiteboard gave me clarity a monitor never did.
- Test on real devices. Emulators lie. Real phones don't.
- Talk to people. Twitter and Product Hunt were priceless for feedback and support.
- Have fun. Seriously.
All of that looked great on paper. The real test was how it would play out in practice, and, much to my surprise, the results were good. Really good.