Back button

Case study: Designing an AI-powered sports app

December 1, 2022
UX design for coaching app for golfers

Overview: Our brief began with the premise that a startup’s best market opportunity was to create a coaching app for the traditional golfing community using custom AI. After an in-depth technology audit and user assessment, we determined that the crux of the opportunity was to leverage platform-based AI technology and target new golfers who are younger, less male, and less white than the traditional golfer.

Our deliverable: A sports coaching application that combines phone-based motion capture with AI-powered recommendations to improve athlete performance.

Detail: We were approached by a team of founders who wanted to create a sports coaching app, specifically a coaching app for the golf industry. The founders had a range of experience and expertise in the golf industry from avid players to semi-professional to lauded coaches. We weren't golfers—this was a new industry for us. But the project came to us because it was a challenging problem of integrating AI and data into a seamless user experience. And that's our specialty.

We started with the data set. One of the founders is an expert golf coach who had built his own software to film golfers in his studio using a complicated, multi-camera setup. The jewel in his software was a representation of the perfect swing as a set of three dimensional coordinates. He could look at a recording and see when a golfer's body moved away from the perfect swing and then provide coaching recommendations for improvements.

Our challenge was replicating his complicated, multi-camera setup and personalized, expert coaching into a phone-based app. How could we get the 3d information we needed without multiple cameras? How could we track a fast-moving swing? And how could we translate a swing into personalized, app-based recommendations?

We started with user research to understand the user and use case. How did people want to be coached? When did they want to practice their swings? What kind of recommendations were they interested in? Which golfers were most likely to use our new app?

What we learned: Despite the stereotype of golfers being older, whiter, and more male, we found that the growth of golf (and importantly the growth of beginners who wanted coaching) was in people who are younger, less white, and less male. Our market research identified an opportunity to stand out in the market with a more modern brand and design.

Next, we worked through the details of the user experience. From onboarding to first trial through to long-term, habit-forming learning. We worked closely with our friends at Noise Studio who are experts in sports app and their insights into the details of how to visualize a golf swing were invaluable.

Finally, we paired our UX design with technical research with our engineering team to figure out how to make it all work. While other startups in the space were developing custom AI for motion capture, we discovered that we could rely on recently developed technology embedded inside iOS. Not only would relying on Apple speed up development, it would also decrease long-term maintenance cost. This allowed us to focus on our own secret sauce: an AI system to identify swing deficiencies, recommend coaching recommendations, and track the success of those recommendations as an internal learning process. We included gamification features to incentivize the user to practice (and help the AI learn).

Check out the project below as screenshots or as a narrated video.

Get in touch if you have ideas for an AI-based application and would like to talk about how we can help.

Onboarding tips for better results

Simple setup and calibration

UX design showing helpful instructions to get the right data

UX design showing motion capture of your swing

UX design showing how data helps you understand your training

UX design showing swing analysis using a 3D visualization model

UX design showing how progress is tracked through swing analysis