Most organizations segment their customers by how often they show up:

“How recently did they log in?”

“How frequently are they learning?”

But what if we’re missing something deeper – and more predictive?

New Research: Why Routines Matter More Than Frequency

A recent study by HBS professor Eva Ascarza and co-authors introduces a powerful distinction between habits and routines.

• Habits are repeated behaviors.
• Routines are repeated behaviors that occur with temporal structure—same time, same context, week over week.

In their research on ridesharing customers, they found that users with structured, time-based routines are:

• More loyal
• More resilient to service failures
• Less price sensitive
• And ultimately, more valuable to the business – even when frequency and spend are the same.

You can read the paper here: “Detecting Routines: Applications to Ridesharing Customer Relationship Management”(JMR, 2024)

But What If Your Product Doesn’t Fit Into a Routine?

I work on a learning product for industry professionals. And it made me wonder:

Do our users even have a learning routine?

Unlike commuting or grocery shopping, professional learning is often problem-driven:

• “I need to figure this out for a project.”
• “My manager asked me to lead this.”
• “I want to sound informed at my next meeting.”

So if structured routines don’t naturally form, does this research still apply?

The Answer: Segment by Triggers and Intent, Not Just Behavior

When natural routines don’t exist, you can still segment by the why behind the behavior.

For example:

• Trigger-based segments: “Learning because of a new project,” “Prompted by a manager,” “Upgrading skills after a role change.”
• Intent-based segments: “Quick-win learners” vs. “Deep divers”
• Engagement patterns: “Burst and drop” vs. “Slow and steady”

This approach gives us a richer lens to understand users – not just how often they engage, but why they show up.

Designing for Routines, Even When They Don’t Exist Yet

Just because routines don’t exist naturally doesn’t mean we can’t design for them.

Here’s the question I’ve been asking myself:

What recurring moments in a user’s workflow or week could learning attach to?

Some possibilities:

• Align nudges with weekly team rituals
• Suggest content tied to recurring tasks (e.g., Monday planning, Friday wrap-up)
• Encourage micro-routines: “10-minute Thursday Think Time” or “Friday Reflection Feed”

These aren’t just reminders—they’re behavioral anchors. Over time, they can evolve into real routines that drive retention and value.

Final Thought: What’s the Shape of Your User’s Time?

Segmenting by behavioral structure – when and why users act – reveals deeper customer value than frequency alone, and offers new design opportunities even when routines don’t yet exist.

So I’ll leave you with this question:

What % of your user value is coming from structured, time-based behavior and what might change if you started designing around it?

I’d love to hear from others working in learning, product, or behavioral design.

Are you thinking about routines in your customer lifecycle?