Ray · AI-driven research

A 20-person study that opened Ray’s first B2B pilot

20 → 14recruited, then analyzed

101workouts in four weeks

1B2B pilot, with GoodPath

Opportunity

GLP-1s are changing how people approach weight loss, and about 11% of U.S. adults are now on one (Gallup, 2026). The medication curbs appetite, but it also comes with fatigue, nausea and muscle loss, and most users aren’t given any guidance on how to stay active.

The goal of this study was to understand how we could better serve GLP-1 users from a check-in perspective, and make workouts more accessible for them.

Key findings

Fatigue was the biggest reason users missed workouts, and how they felt each week depended a lot on when they took their shot. Only 2 of 14 had been told by their doctor to strength train, and many turned to walking on harder days. What kept everyone going was the group chat.

These insights shaped the product direction for Ray’s first B2B pilot with GoodPath.

Process

A four-week study

I recruited 20 participants on GLP-1s and had them use Ray for four weeks. Each week they did two workouts with a quick check-in before each one, plus a deeper weekly check-in, and everyone was in a group chat together. I cast a wide net when recruiting, so 13 of the 14 participants I analyzed were new to Ray. At the end, I ran exit interviews with 7 of them.

View participant guide

Recruit

Screener surveyFlyers, Facebook groups, Reddit, Slack, Craigslist and Ray users

Four weeks with Ray

Week 1Get to know Ray
Week 2Talk to Ray
Week 3Make progress
Week 4Push furtherVideo calls

Exit

7exit interviews

Every week

Thursday · 30 minThe Quick BurnStart

2 workouts a week

Ready for today?

12345

FatigueLow energy

A quick check-in before each workout

Your shot week

MTWTFSS

Day 3 after your shot · same dose

A deeper weekly check-in on their shot cycle

MSwapped to a 20 min walk today

DSame! Friday is my big day 💪

A group chat to keep each other going

SSaw your energy was low. Want a lighter session Thursday?

Yes please 🙏

A text from me when the agent flagged someone

Flag, then a human reaches out

I built an agent in Claude that read every check-in and flagged low readiness, fatigue, dose changes and missed check-ins. It also caught two or three participants whose activity didn’t add up. Then I followed up manually to better understand their situation.

Check-ins come inBefore each workout, on how they feel that day. Each week, a deeper look at their shot cycle and progress.

An agent flagsReadiness below 3 of 5, fatigue, low energy, missed check-ins, unusual activity.

I follow upA personal text to understand their situation. Video calls came only in the last week.

The plan adjustsLike timing workouts around the shot, which made working out possible again for one participant.

The study dashboard: check-ins today, WhatsApp check-ins to answer, workouts and walks this week, a Reach out today list flagged by the agent, the most active participants, a 14-day check-in grid and this week’s flags. Names are invented. — The dashboard I used to run the study. Names are invented to protect participants.

What we learned

Four learnings from the check-ins and conversations, from my readout to the Ray team.

Learning 01

Fatigue costs the most workouts

~30%of workout check-ins mentioned fatigue

65% of weekly check-ins rated workouts “tough but worth it,” and users asked for recovery and nutrition tips.

Learning 02

The shot sets the shape of the week

15%of check-ins involved a dose change

“I take my shot on Mondays, and Tuesday to Thursday I experience a lot of fatigue and muscle aches.” Backloading the week worked for two participants.

Learning 03

Knowledge about GLP-1 varies

2 of 14were told by their doctor to strength train

3 of 14 wanted to know why an exercise was in their plan, and Ray felt less encouraging after week 2.

Learning 04

Users walk a lot

174walks logged by 10 participants

Walks didn’t count toward the plan, and walking replaced workouts that didn’t feel doable that day.

What kept users going

The group was the accountability

Participants loved checking in alongside each other in the group chat. That shared accountability kept them showing up, and it shaped how we improved Ray’s own check-in experience.