Les Mills: From batch emails to preventing churn before it happens 

When COVID-19 hit in 2020, Les Mills Plus saw something remarkable: 900% subscriber growth as people abandoned gyms for home workouts. But having all those new subscribers meant nothing if they couldn't communicate with them effectively.

Mindy, who leads global digital engagement strategy for Les Mills, found herself managing a massive influx of movers (Les Mills' term for their workout enthusiasts) with essentially one communication strategy: batch and blast everyone with the same message. In our recent "Workflows That Workout" webinar, she shared how her team evolved from that basic approach to building a sophisticated churn prediction system that retains 53% of at-risk subscribers.

Before Les Mills could level up, they needed to solve a fundamental problem. Every workflow change required engineering support. Every new segment demanded developer time. The marketing team had ideas but lacked the ability to execute them quickly.

Moving to Customer.io changed that dynamic immediately. The drag-and-drop workflow builder meant Mindy could spin up campaigns, test messaging variations, and adjust targeting without submitting engineering tickets. This wasn't just about convenience. It was about being able to respond to user behavior in real time rather than waiting weeks for developer availability.

But self-service tooling only matters if you have the right data to work with. Les Mills set up an onboarding survey that captured crucial information about each new subscriber:

• Their fitness experience level
• Their workout preferences
• Their goals

That survey data flowed directly into Customer.io through a webhook, creating the foundation for personalized messaging from day one.

The result was a branching onboarding flow where beginners got more guidance and motivation while experienced exercisers were directed straight to the workout types they cared about. Same tool, completely different experiences based on what each person actually needed.

As Les Mills added more features to its platform, particularly its science-backed fitness plans and challenges, the team hit a new wall. They were tracking plan signups as user attributes, but attributes have limitations. They update asynchronously and don't capture the timing of actions particularly well. And when subscribers could only be in one plan at a time, the data got messy fast.

The breakthrough came from rethinking how to structure their data. Instead of updating profile attributes whenever someone joined or left a plan, Les Mills started sending events for these actions. When a subscriber signed up for the "Getting Started with Pilates" plan, that became an event that triggered a workflow immediately. When they completed their first workout in that plan, another event occurred. When they stopped following the plan, another event.

This shift from attributes to events unlocked real-time intervention. If someone signed up for a four-week strength challenge and then didn't complete a workout for five days, the system could catch that immediately and send a re-engagement message. The timing became surgical rather than approximate.

Most marketing teams track opens and clicks. Fewer track conversions. Les Mills realized they needed to close the loop between their messaging and actual workout behavior.

They started passing workout completion events into Customer.io and setting those as conversion goals within their campaigns. Now they could answer questions that actually mattered: Did people who received this onboarding email series complete more workouts than those who didn't? Which plan reminder messages drove the most workout completions? What messaging cadence kept people most engaged?

This also meant they could stop exporting data to external reporting tools for every analysis. The tagging system in Customer.io let Mindy label all plan-related campaigns with a single tag, then pull up comparative performance across all of them in seconds. No data warehouse queries required.

The shift to conversion tracking created another benefit: holdout testing. By randomly excluding a small percentage of subscribers from campaigns and comparing their behavior to those who received messages, Les Mills could prove their messaging was actually working. This wasn't just good for optimization. It helped justify continued investment in lifecycle marketing to leadership.

Everything Les Mills built over three years came together in their churn prediction model.

Their data engineering team created a machine learning algorithm that scored subscribers based on behavioral patterns: subscription tenure, workout frequency, engagement with different features, and dozens of other signals. The model tagged movers as high-risk or low-risk for churning.

That risk score fed into Customer.io as a profile attribute. But the real power came from combining it with event-based triggering. The system could now identify someone who was both high-risk for churn and approaching their renewal date, then intervene at exactly the right moment with exactly the right offer.

Les Mills offered discounted annual plans to monthly subscribers showing churn signals. For subscribers who had stopped working out but weren't near renewal, they sent re-engagement campaigns highlighting new content in their preferred workout categories. The messaging varied based on churn risk and context.

The results were dramatic. Within two months of launching the churn model, Les Mills retained 53% of subscribers who were predicted to churn. They converted 80% of at-risk movers to annual subscriptions, locking in revenue and dramatically improving customer lifetime value.

The program worked so well that the Les Mills team was named a runner-up for a New Zealand national marketing award.

Most companies don't have a data science team ready to build churn prediction models. But the principles that made Les Mills Plus successful apply at any stage of sophistication.

Start by removing bottlenecks.

If your team needs engineering support for every campaign change, you'll never iterate fast enough to find what works. Self-service marketing tools matter, but only if you actually use them to experiment and learn.

Structure your data for the questions you want to answer.

Les Mills moved from attributes to events because they needed to understand timing and sequence, not just states. Think about what you're trying to learn from your campaigns, then make sure your data structure supports those insights.

Track conversions, not just engagement.

Opens and clicks tell you if people saw your message. Conversions tell you if your message actually changed behavior. Even if you're just tracking one conversion event to start, that's infinitely more valuable than optimizing for opens alone.

Build feedback loops.

Every campaign Les Mills ran taught them something about their subscribers. Survey responses improved targeting. Workout completion data informed future messaging. Churn signals triggered interventions. Your lifecycle marketing system should get smarter over time, not just send more messages.

Justify your work with holdout testing.

When you can prove that subscribers who receive your campaigns behave differently than those who don't, you move from "marketing is probably helping" to "marketing drives X% improvement in retention." That clarity matters when budgets get tight.

Les Mills didn't build their lifecycle program overnight. They started with basic onboarding emails, discovered limitations, addressed them systematically, and kept building on what worked. Three years later, they're preventing churn before it happens and proving ROI with data.

You can watch the full "Workflows That Workout" webinar with Les Mills to see Mindy walk through her actual workflows and share more details about how the team structures their campaigns.

The path from pandemic-era chaos to predictive retention took three years of steady iteration. If you’re curious how your team can get started with your own successful lifecycle campaigns, set up a walkthrough of Customer.io today.