How a Full-Time Teacher Built a Nearly $1M App Business

After years of building an app for a problem users didn't feel, the founder refused to put life on hold for a big build. He shipped Shot Pattern three weeks after starting development (2023), using ChatGPT copy/paste to help with code he couldn't figure out alone. The app had only two features, neither offered by other golf apps (seeing strategy in real time by measuring sideways on the hole, a portable version of the Google Earth planning method golf strategy thought leaders already taught), and none of the features people expect from a golf app. The target audience was already primed by those thought leaders, so no convincing was needed.

In the first summer the only marketing was the founder sending out tweets and jumping into golf strategy conversations on Twitter/X. People who saw it signed up readily and sent direct messages praising the app and asking for more features.

After buying a $5,000 golf course data API, the founder found it enabled a new feature: simulating 1,000 shots on a hole (Mark Broadie's method) using the golfer's own data and the course they're traveling to. Halfway through building it, he posted a teaser tweet on X saying the feature was coming, with a cheeky hook that it 'would be highly illegal' to use during play. The audience already believed in this way of thinking about golf.

When a company offered to buy the app at ~$2K lifetime revenue and asked for a deck of tech stack and user metrics, the founder (finance background) saw that a multiple of revenue or profit would give a tiny valuation. He pulled the RevenueCat State of Subscription Apps benchmarks and compared his RevenueCat numbers: trial start rate and trial conversion rate were both 99th percentile, download-to-paid also very high. He concluded the small dollar amounts hid a strongly converting funnel and declined the $75K offer, despite the buyer threatening to build the feature themselves.

At $39.99/year he believed paid ads couldn't work, based on a Sub Club chat rule that you need at least $75/year. He looked at competitor pricing (top of the golf GPS market at $99/year) and chose ~$75 as a middle-to-top price, since the app still felt unfinished. Then he asked which features would make $75 feel fair to his target users, and with limited time (newborn, full-time job) built only the most valuable next step: shot tracking features, including recording demo videos at 4:30am.

Before spending on Meta he already knew from RevenueCat that trial start rate, trial conversion and conversion to paid were all strong (99th percentile vs benchmarks), so ads were 'fuel on the fire'. He says you can't jump to paid ads as the secret to success; you need data saying the app converts first, then get more people to see it.

Once shot tracking was live, he set up a Meta campaign optimized for installs (trial start events were not wired up yet). He'd heard an ad set needs at least three creatives and that under $10,000/month spend won't exit learning, so he budgeted about $300/day using the few thousand dollars he had. Creatives: a UGC creator 'play this hole with me' video with screen overlay, a paid video from TikTok creator Speedgolf Rob, and a 30-60 minute screenshot voiceover he made in his basement. Meta put most spend on the basement video. He watched the dashboard daily and was ready to turn it off the moment it stopped working.

Made in 30-60 minutes in his basement: a screenshot/screen-recording voiceover tutorial, with the QuickTime recording dot still visible, thumps and clicks audible and a misspoken number left in. Hook: 'most golfers could use a little help understanding strokes gained' (confusing to some, a chance for experts to feel knowledgeable). It then walks through a real 430-yard par 4 showing strokes gained for a drive down the middle, into the rough and into a bunker, and ends 'try out Shot Pattern so you can see strokes gained before you hit the shot'. It showed the core audience exactly what they already wanted to do.

With ~$1.50 CAC returning $4-5 per customer (about $4 per $1 within 8 days), he planned a bank line of credit as a cushion until Apple payouts arrived, and ended up using an interest-free credit card, spending as much as he could while the math held. Because golf is seasonal, he pushed hard in July and August, then turned ads off at the end of the season.

Ongoing process: build features centered on the ideal customer (often what users ask for), then grab a recently built feature, go with the first idea that comes to mind, clip it up and publish. Deliberately make it look unlike the other ads the audience sees. Run them in a Meta ad set at about $500/day and let Meta pick the best converters.

Found TikTok golf creator Speedgolf Rob (hundreds of thousands of followers, featured in Golf Digest, eye-catching style, high engagement) after his videos kept appearing in the founder's feed. Rob offered: pay for him to play in a golf tournament and he makes an ad about using the app to prepare. The video ran as one of three creatives in the Meta install campaign. A separate UGC creator also made a 'play this hole with me' video for the app.

Before video, he tried static ads first.

Alongside Meta, he advertised on podcasts and pays creators and influencers to share the app, cross-posting much of it on his own page. He gave the app to a YouTuber and now pays him to embed content in his videos.

With a contractor, ran three paywall experiments in a row, each shorter than the last, removing words until the paywall felt like almost nothing. Each was judged on trial start rate.

Against the contractor's advice, rebuilt a long paywall like the original: an image, a long list of features with descriptions, an extra line of social proof, and a 6-month package option for price anchoring. Goal: a user who sees the one feature they want plus lots more feels it's a great deal. Measured on revenue rather than trial start rate, against the short winning paywall.

Reviewed many apps' onboarding screens and built one like everyone else's: social proof, ATT prompt, permission requests, feature hints and personalized setup, screen after screen.

Asked whether onboarding was too long or too boring, and instead of more screens built an immersive interactive tutorial: the user interacts with the onboarding, learns how the app works and how to think the way the app needs, before entering the app.

Noticed a big drop-off where the tutorial switched to a second situation (tutorial part two). Ran an experiment removing part two to shorten onboarding, measured in Firebase A/B testing on sign-up rate and trial conversion.

Over the summer, the solo founder hired a contractor with extensive golf industry and product management experience (outside apps) to build an experimental mindset and look for easy gains. They worked together over Slack on paywall and onboarding tests, with the contractor pushing simpler designs and the founder overriding him when he disagreed.

The app is intentionally not for all golfers: it targets competitive golfers. When a casual golfer finds onboarding too advanced and leaves, he lets them go rather than simplifying, because making it not for them makes it more for his people. As a one-person side business he can afford a small TAM.

Golf GPS is dominated by some of the oldest, highest-revenue apps (e.g. Golfshot, 10+ years). People searching generic terms are not his audience, who hang out in other places, so he chose paid ads over ASO.

As a coach he saw athletes' pressure and identity struggles firsthand and built a meditation/resilience app for high school athletes, working on it for four to five years.