top of page

CONSUMER COACHING PLATFORM

Increasing retention with personalized Gen AI lesson plans

PRODUCT DESIGN

RETENTION STRATEGY

MOBILE APP

GEN AI

Swingtweaks 1.png
Role

Lead Product Designer

Responsibilities

End-to-end UX & UI Design Process 

Team

Product Manger 

Engineering

Timeline

Q1 2026

Tools

Figma

Claude

AT A GLANCE

Overview

SwingTweaks is a mobile golf coaching platform connecting everyday golfers with PGA professionals for personalized swing analysis. As Lead Product Designer, I led an end-to-end redesign effort focused on lesson retention. First-purchase conversion was strong at 25%, but only 20% of golfers returned for a second lesson. Partnering with product and engineering, I used hypothesis-driven research to identify post-lesson uncertainty as the core retention barrier, then designed an AI-powered Practice Plan feature that turned coaching feedback into structured weekly training, creating a clear, motivating path back to the next lesson.

Goals
  • Replace assumptions about churn with evidence from real user research

  • Give golfers clear direction on what to do between lessons, not just after one

  • Use AI to connect a coach's personalized video analysis to specific drills and sessions, turning feedback into actionable practice instead of a one-time video a golfer watches once and forgets

  • Design a scalable solution that doesn't multiply coaching costs per golfer

The Challenge 

BUSINESS NEEDS

Grow "tweaks per tweaker per year." Golfers averaged just 1.2 lessons annually, meaning most never returned after a first session.

USER NEEDS
Golfers needed a clear sense of what to practice and when to come back. Without it, even a positive lesson experience faded into inaction.
PROBLEM TO SOLVE

Golfers who complete a lesson often don't know what to do next, leaving them without a clear reason, or a clear plan, to return.

RESEARCH

Gathering Insights

While it would have been easy to assume users weren't satisfied with their lessons, we wanted to understand the root cause before jumping to solutions. We needed to move from assumptions to evidence.

Two Cohorts Interviewed:

REPEAT CUSTOMERS

Highly motivated, self-directed, actively practiced between lessons, sought ongoing feedback.

ONE-TIME CUSTOMERS

Satisfied with the lesson, but uncertain what to practice next, unsure when to return, no structure between sessions.

Key Themes and Insights

These insights drove the prioritization of features and ensured alignment of goals.

DIRECTION,
NOT MOTIVATION

Golfers who churned weren't unmotivated. They were uncertain what to do next.

STRUCTURE BUILDS CONTINUITY

Golfers who did self-directed practice between lessons correlated with repeat bookings.

VALUE
PERCEPTION

Some churn wasn't about experience quality. Rather golfers believed the original problem was already solved.

HOW MIGHT WE

How might we give golfers a clear, structured path between lessons, so uncertainty doesn't become the reason they don't return?

IDEATION

Hypothesis, Tradeoffs & Design Focus

While it would have been easy to assume users weren't satisfied with their lessons, we wanted to understand the root cause before jumping to solutions. We needed to move from assumptions to evidence.

HYPOTHESIS ONE

A clear call to action for when to return would increase repeat purchases. 

HYPOTHESIS TWO

Structured practice tasks would remove ambiguity between lessons.

Design Focus

To meet tight deadlines, we prioritized rapid iteration, starting with highfidelity mockups with a priority focus.

Swingtweaks home and tweak.png
  • Replace assumptions about churn with evidence from real user research​

  • Turn coaching feedback into clear, actionable next steps

  • Keep golfers motivated between lessons

  • Reduce cognitive load during practice

  • Create a natural return point for additional coaching

  • Deliver a consistent experience across different coaching styles

  • Use AI to connect personalized coach feedback to relevant drills and practice sessions

  • Design a scalable solution without increasing coaching costs per golfer

Evaluating Interventions

We explored several approaches, balancing impact with effort and business constraints. The Practice Plan emerged as the strongest opportunity. It directly addressed the core problem of uncertainty after the lesson ended.

​

The catch! Manually building a personalized plan for every golfer would significantly raise coaching costs. We partnered with engineering to solve this with an AI-assisted approach instead.

Evaluating Interventions

We explored several approaches, balancing impact with effort and business constraints. The Practice Plan emerged as the strongest opportunity. It directly addressed the core problem of uncertainty after the lesson ended.

​

The catch! Manually building a personalized plan for every golfer would significantly raise coaching costs. We partnered with engineering to solve this with an AI-assisted approach instead.

HOW IT WORKS
  1. The golfer receives a personalized video analysis from their PGA pro after they submit their "Swingtweak"

  2. An LLM analyzes that feedback and connects it to specific drills and sessions

  3. A personalized 3-week practice plan is generated, linking the coach's actual feedback to concrete practice

  4. The golfer receives structured weekly exercises and a clear finish line: a defined point of completion that prompts them to book their next lesson.

Swingtweaks practice plans 2.png

Trade-offs

Key decisions we made to balance user needs, business goals, and delivery constraints.

SCALE VS COST

Saclable AI-generated plans vs. the cost of creating plans manually

CLARITY VS FELXIBLITY

A clear, fixed 3-week structure vs. a more flexible, adaptive experience

SPEED VS DEPTH

Faster feature delivery vs. more extensive user validation

USER TESTING

Testing The Solution

While it would have been easy to assume users weren't satisfied with their lessons, we wanted to understand the root cause before jumping to solutions. We needed to move from assumptions to evidence.

FINAL DESIGN

The MVP Design

While it would have been easy to assume users weren't satisfied with their lessons, we wanted to understand the root cause before jumping to solutions. We needed to move from assumptions to evidence.

Homepage Key Considerations
SEE WHAT'S NEXT
See what to practice next in your lesson plan
QUICK ACCESS
Drills, coach feedback and recent tweaks in one place
LESS FRICTION
Fewer steps between feedback and practice
SUBMIT YOUR SWING
A clear path back to coaching
Swingtweaks 8.png
Lesson Plan Key Considerations
ACTIONABLE FEEDBACK
Turn coaching feedback into clear next steps
BETWEEN-LESSON MOTIVATION
Keep golfers engaged and practicing between lessons
CONSISTENT EXPERIENCE
Create a reliable experience across coaching styles
SCALABLE
AI supports personalized practice without increasing coaching costs
Swingtweaks practice plans.png
Swingtweaks 7.png
Swingtweaks 4.png

SUCCESS METRICS

Measurable Impact

+40% Engagement
Increase in repeat lesson engagement
25% Increase In Sales
Tweaks sold per golfer per year
2 User Cohorts
Interviewed to identify retention barriers

PROCESS & APPROACH

Process & Approach

Evaluating Interventions

3. Evaluating Interventions

Rather than jumping directly into a solution, we explored several approaches and weighed their potential impact against implementation effort and business constraints.

WHAT WAS THE OUTCOME?

The Solution

Research revealed that many golfers weren't failing to improve because they lacked motivation. They were struggling because they lacked direction. To address this, we introduced Practice Plans, an AI-powered feature that transformed coaching feedback into structured weekly training plans. The goal was to remove ambiguity between lessons and create a stronger connection between coaching sessions.

By converting a completed lesson into actionable next steps, golfers received:

  • Clear practice goals

  • Weekly training activities

  • A structured learning path

  • Defined milestones for returning to coaching

A LOOK TO THE FUTURE

Next Steps

Given the time constraints, our team delivered an exceptional product. However, if the engagement were extended, my approach would focus on further validation, refinement, and iteration to maximize usability and adoption. Ensuring the product is not only functional but also optimized for real-world use, driving long-term value and business impact.

Lessons Learned

Users weren't asking for more coaching. They were asking for more direction. By looking beyond acquisition metrics and focusing on retention, we uncovered a gap between receiving feedback and knowing how to act on it. The resulting Practice Plan feature increased repeat lesson engagement, validating our hypothesis, but also revealed that improving a metric and solving a business challenge are two different milestones. The work succeeded, but it also clarified what problems still remained to be solved.

What's Next

The Practice Plan feature successfully increased lessons per golfer from 1.2 to 1.5, validating our hypothesis that golfers needed more structure and guidance between coaching sessions. While the results were encouraging, they also revealed that improving retention would require more than a single intervention.

​

Future exploration would focus on helping golfers build lasting practice habits through progress tracking, personalized coaching journeys, and stronger accountability mechanisms. The next challenge wasn't helping users understand what to do next, but helping them consistently follow through and return as part of an ongoing improvement journey.

© 2026 by Aja Deren. All rights reserved.

  • LinkedIn

Powered by early mornings & sweet treats 

bottom of page