Tempo Town

Golf, before you're good at golf.

An AI swing-analysis concept for the beginner golfers the market ignores. Fundamentals first, mechanics second, and a coach marketplace that only kicks in when you're ready for it.


Role:

Senior Product Designer

Tools:

Figma

Domain:

Consumer · Sports · AI

1

Weekend from concept to prototype

3

Competitors studied — Sportsbox AI, V1 Golf, Hackmotion

25M

US golfers. Beginner and recreational players are the largest segment.

The Spark

I spent this year trying to teach myself golf.

Every weekend at the range, I'd film my swing on my phone and scrub through the footage in slow motion, trying to figure out what I was doing wrong. No coach, no lesson plan, no data — just me, a tripod, and the Instagram slow-mo feature.

Most of the time I couldn't tell what was actually wrong. I knew something was off, but I didn't have the vocabulary for it. The obvious answer was to hire a coach. The problem: coaches are expensive, and I wasn't yet good enough for a coach to help me.

The gap between "I don't have a swing yet" and "I can work with a coach" had no product filling it. That's where Tempo Town started.

The Gap

The AI golf space is built for golfers who already have a swing

A market scan of the four tools most serious golfers mention:

Sportsbox AI

Application

3D biomechanics from a single phone video. Markerless, sensor-free. Positioned around coaches and tour-level players — the pro app requires a coach login.

V1 Golf

Application

Video analysis with drawing tools, ball tracing, and skeletal tracking. Core distribution is the coach-student lesson workflow — students send swings, instructors send back annotated lessons.

Hackmotion

Hardware + App

Wrist sensor measuring flexion, extension, and rotation. Its own App Store description states the product is "intended for golf coaches and advanced golf players."

All three assume you already have a swing worth analyzing. None meet a beginner where they actually start: grip, stance, ball position, alignment. The beginner — the most frustrated, most likely to quit within two years, and by far the largest segment of the market — has no serious tool designed for them.

The Thesis

Fundamentals first. Mechanics second. Coaching as a graduation path, not an upsell.

Tempo Town inverts the assumption baked into every existing tool. Start with what beginners actually need, let the coach marketplace activate only when the user has built a baseline worth coaching. No one sells you a lesson you're not ready for.

01

Onboarding in 3 screens, not 5

Cut typical miss and handedness from onboarding. Handedness can be inferred from the first swing video. Typical miss is meaningless until there's baseline data. Every pre-upload screen is a drop-off point.

03

Analysis output built for non-experts

Every metric shows a 0–100 scale, a skill-tier benchmark, and a plain-English interpretation tied to the specific score. A number without context doesn't teach, it intimidates.

02

Two entry paths, explicitly labeled

"Still learning how to hold the club" vs. "I have a repeatable swing." The app doesn't pretend one experience serves everyone. Beginners route to fundamentals; improving players route to mechanics.

04

Coach marketplace gated by baseline

Coach cards stay locked until the user hits a minimum threshold. The copy makes the thesis visible: most golfers get the most out of coaching once they have a repeatable swing to work from.

What I’d Validate Next

Three bets worth testing.

1.

Does the fundamentals-first routing actually reduce beginner churn?

The core bet. Would need two cohorts — one starting with grip/stance checks, one starting with swing analysis — and a 4-week retention comparison.

2.

Is "gated coach marketplace" a feature or a friction?

Users may resent being told they're not ready. Test whether framing it as a milestone ("you're coach-ready") outperforms framing it as a gate ("not yet").

3.

Can phone-video AI actually diagnose fundamentals?

The whole concept depends on AI being able to diagnose grip and stance from a phone video. I haven't validated this. A real build would need computer-vision feasibility testing before design.