An in-app chat feature had been developed with the hope of fostering new connections among existing WW members. Three different use-case pilots had failed to gain traction. Due to organizational changes, the feature was about to be shelved. Executive leadership turned to the in-person End-to-End experience team for a Hail Mary.
Ideally, new products are developed from a place of insight or need. Not every project is ideal. Some of the fun of design is finding resonance in messy spaces.
Approach
With executive pressure to launch immediately, I laid out a phased strategy instead: building on existing research rather than starting from scratch, piloting in smaller markets, and developing a change management approach to recognize the impact Chat would have in the WW ecosystem.
Key insights from prior pilots and the behavioral science team shaped our direction:
Workshop attendance hadn't recovered post-Covid. Lack of community connection was the most cited reason.
Members within workshops already know each other, removing the "stranger danger" friction from previous pilots.
Members are initially attracted by the Coach, but return over time for the community.
We designed a Chat experience structured around Coaches, WW's most defining differentiators, grouping members by the Coach they most frequent.
Navigating competing priorities
Weeks before our first pilot, product partners informed us that Chat's entry point would be moved deep in the app's hamburger nav, 3-4 taps away. In failed prior pilots, ~70% of participants cited lack of awareness as their reason for not engaging. Buried access would have killed the project.
I worked with the executive team to negotiate a compromise: pilot participants would retain 1-tap access via the Home Screen, with the new access point integrated into the new design system later in the year, which meant accepting a delayed national launch.
Preferring the right member experience over speed, I made the call to take the compromise and push back the timeline.
Refinement
After two pilots in different markets, key insights challenged our initial assumptions, particularly around the Coach's role. While groups were Coach-organized, they didn't need to be Coach-led. The most engaged groups had Coaches who participated as enthusiastic members rather than as "coaches."
Optimizing Coach time in Chat, without compromising member experience, saved over $1M in staff hours.
Working with Field Operations, we crafted a Coach playbook and training to help coaches develop their own voice while identifying new ways to incorporate Chat into their Workshops.
Launch + Scale
Coach-organized chat groups created
Premium and/or Lifetime members assigned
Week 1
Week 6
After 4 months, daily active participation sustained at 1-2%.
Voices from the community
The hero brand hamburger rolls are 2 pts and taste like regular bread... found at ShopRite in the bakery area
I liked Coach BJ's quote, "WW gives you tools, not rules." Really good workshop. So many good comments, tips, and reminders.
Hey Lani this is Betsey! Totally agree, a lot of inspiration from our "peeps"!
Very emotional meeting yesterday. Very good as usual. This is Laura Doty
This chat is SO cool! I haven't been to a meeting or tracked in weeks, feeling the hopeless emotions that can creep up on us along this journey. I am dedicating to attending the meeting on Saturday because you all inspire me and remind me I'm not alone.
Lauren8015, WW Member
Impact
Chat's full business impact won't be realized for at least a year. Formal experiments to establish causation between Chat activity and KPIs were being developed before WW decided to slim down its headcount as well.
Early Correlation
Pilot locations showed a correlation between Chat group activity and both Workshop attendance and app engagement.
Staff Efficiency
Reframing the Coach role saved over $1M in staff hours while maintaining or improving the member experience.
Community Health
Coaches continue to share stories of unexpected ripple effects. Members reconnecting. Motivation sustaining between meetings.
Ongoing Development
We implemented a learning plan to measure ongoing impact and guide future iterations of the feature.
Nobody does amazing work alone.
This project was made all the better with: