YouTube | Spotify | Apple Podcasts
Why Gym Owners Should Stop Building Custom AI Apps
Dan Cotton, host of the Cheer Biz Podcast and self-proclaimed AI super fan, makes the case for why most cheer gym owners should resist the urge to build their own custom apps with AI, and explains what they should focus on instead.
An AI Fan’s Surprising Warning
Dan Cotton opened this episode of the Cheer Biz Podcast with a confession: he is, in his own words, a “super duper AI fan.” He and Danielle even started a Skool group dedicated to teaching gym owners how to use AI in their businesses. But despite that enthusiasm, Dan spent this episode making an argument that might surprise longtime listeners: most gym owners should not be spending their time right now building their own custom app with AI.
Before diving into the episode, Dan reminded listeners to join the Cheer Gym Owners and All-Star Cheer Coaches and Owners groups on Facebook, and pointed anyone curious about applying AI in their gym toward the AI for Cheer Gym Owners Lab on Skool.com, a low-cost monthly community where he and his team share what they are learning. He also flagged the upcoming Cheer Biz Accelerator events in Warrensburg and Portland this October and November, with two more stops planned for January and February, where gym owners can get hands-on help scaling their businesses.
Two Attempts at Building an App
Dan’s caution comes from experience, not theory. About two years ago, he discovered a platform called Replit that let him build software through “vibe coding” — describing what he wanted in plain language and letting AI generate the code. He started small, digitizing the way DreamCamps tracked its camp games and determined camp champs, a process he had previously managed through spreadsheets and handwritten notes.
That small project grew. Dan kept adding features, including a system for athletes to book private lessons with coaches during individual camp, replacing a nightly routine of printing out sign-up sheets. But the more he bolted onto the app, the more bloated and fragile it became. When he rolled it out at camp and told everyone to book their private lessons through it, the app broke. He estimates he put more than 100 hours into building, fixing, and adjusting it, and it still didn’t work reliably.
This year, armed with lessons learned and new tools, Dan rebuilt the app from scratch on a different platform. He started in December and launched in June, spending roughly 80 hours on it over that stretch, alongside his other work. This time, it performed beautifully all summer. Dan was quick to point out why the second attempt succeeded where the first struggled: he now has head instructors and an operations director running day-to-day delivery at DreamCamps, which freed him up to spend that kind of time building. Most gym owners, he noted, don’t have that luxury.
Know Your Numbers First
Dan’s first reason gym owners shouldn’t be building custom apps is a gut check on priorities. He posed a direct challenge: if someone asked for your gross revenue, net profit, gross profit, customer acquisition cost, lifetime value, cash on hand, marketing strategy, or org chart, could you produce accurate answers within about an hour?
If the answer is no, Dan argued, you have far more pressing things to work on than a custom app that solves one narrow problem. He described this pattern as a common trap: a gym owner identifies a single pain point, decides to try building a solution with AI, and slides down a rabbit hole. It feels productive. People are impressed when they see the finished product. But if it isn’t improving the core metrics and delivery of the business, Dan called it “an error disguised as momentum” — activity that feels like progress without actually moving the needle. His advice is to nail down the fundamentals of the business first, and only then consider more complex technical projects.
AI Should Save You Time, Not Cost It
The second reason ties directly to how gym owners are already using AI. Dan explained that if someone has only ever used AI as a more advanced Google search, that’s a signal they probably shouldn’t be the one building a custom app with it. In his view, AI’s real value for gym owners is automating, streamlining, and organizing — clearing out the inbox, running marketing tasks, and tracking the numbers that matter, not generating brand-new software from scratch.
He broke down the real cost of building an app in concrete terms. Even the simplest possible app takes about two to three hours to produce a baseline draft, and five to ten hours to reach something that basically works, once you factor in logins, hosting, and setup. The more ambitious builds Dan has worked on personally have run 60, 100, even 200-plus hours.
To put that in perspective, Dan ran the math on 60 hours of work valued at $20 an hour: about $1,200 worth of time. And that figure doesn’t include token costs, hosting fees, or the complexity of setting up authentication and integrations correctly. His question to listeners: is your time really worth only $20 an hour? For most gym owners, he suspects the honest answer is no — and that’s before accounting for what else that time could have gone toward.
The Hidden Work of Iteration and Security
Dan also warned that AI-built apps are rarely as finished as they first appear. He described the common experience of getting an impressively fast first draft from AI, feeling like the project is done, and then discovering — sometimes a day later — that key pieces don’t actually work the way they were supposed to. Building anything beyond the simplest tool means arguing with the AI, correcting it, and cycling through rounds of testing, breaking, and fixing.
Security is an even bigger hidden cost, especially for gyms handling athlete data like names and dates of birth. Dan cautioned that standard AI coding tools — he named Codex, ChatGPT, and Claude specifically — cut corners on security by default. Doing it properly means running additional security checks, reviewing row-level security, and generally treating the build with the same rigor a professional developer would. It’s possible to do, Dan said, but doing it right is a genuinely complex undertaking, not a weekend project.
When It’s Worth the Gamble
Dan didn’t say building a custom app is always the wrong move. He compared it to gambling: never risk time you can’t afford to lose, just as you’d never bet money you can’t afford to lose. If a gym owner is technologically inclined and genuinely has the spare time, he encouraged them to experiment and see how far they can take it — as long as they can accept the possibility that the project fails, or that the knowledge gained along the way is worth the time spent regardless of outcome.
But for most gym owners, Dan’s advice was to wait. He drew a comparison to operating systems: technically, anyone could build their own with AI, and some dedicated hobbyists do, but almost nobody does because it would be a poor use of their time when a great option already exists off the shelf. He hinted that his team is actively building products specifically designed for cheer gyms and cheer gym owners, and offered to do the hard work of iterating with AI tools so gym owners don’t have to. His closing suggestion: invest your time in learning how to use AI to automate and streamline what you’re already doing, rather than spending it trying to build something entirely new from the ground up.
