Building a SaaS Platform and Mobile Apps as a One-Person Shop: Lessons From PickCommish

I build software for clients. PickCommish is software I built for myself, which means I paid for every mistake personally. Here's what it taught me, and why it matters if you're thinking about custom software for your own business.

Start with a problem you actually have#

I run football squares boards for my groups every season. Doing it by spreadsheet, text message and memory is miserable: who paid, who picked which square, who won which quarter, who's owed what.

That pain is why PickCommish exists. Every early feature came from a real annoyance, not a brainstorm. The best custom software projects I've done for clients start the same way: with the thing that eats hours every week.

The stack, in plain terms#

• Web platform: Laravel 12 on the back end, React on the front end, connected with Inertia so it feels like an app without the overhead of two separate systems.
• Mobile apps: React Native with Expo, published on both the App Store and Google Play.
• Payments: Stripe, including subscriptions and a coupon and promo code system that stays in sync with Stripe.
• Sign-in: Google and Facebook login alongside email, because every extra password is a reason for someone to quit signing up.

The web and mobile apps share one back end. One source of truth, one set of business rules, one place to fix bugs.

Features that came from real use#

• Live squares tracker that updates as games are played, so nobody's refreshing a spreadsheet.
• Race View standings for pools that run all season.
• Draft order randomizer, because "who picks first" is its own argument.
• Geoblocking for states where this kind of pool needs extra caution, built in as a compliance measure rather than bolted on later.

Then it got hardware#

The PickCommish Pug is a small custom device built on an ESP32 microcontroller with LEDs, audio and a tiny screen. It connects to the platform's API and celebrates when your square hits. It's partly a fun product and partly proof that the platform's API is solid enough to build on.

Getting there meant debugging Wi-Fi connections, time syncing, an audio module with some undocumented quirks, and LED timing. It's the same kind of problem solving that shows up in client work, just with more blinking lights.

Lessons that apply to any business software project#

Payments and compliance go in early. Retrofitting billing or legal restrictions into a live system is painful and risky. Design for them from day one.

One back end, many front ends. If you'll ever need a mobile app, a customer portal and an internal dashboard, build a single core system they all talk to.

Marketing is part of the build. Landing pages for search, clear onboarding, and ads testing were as much work as some features. Software nobody finds doesn't help anyone.

Ship small and often. Most improvements to PickCommish went out as small releases during the season, based on what users actually did. Big-bang launches hide problems.

Own your stack. Because I built it, I can change anything. That's the real advantage of custom software over bending an off-the-shelf tool that almost fits.

Have a process that runs on spreadsheets and group texts? Let's talk about it.