Back
Case Study 01 · Marketing Analytics

How smart-device habits reveal who to acquire.

Analyzing a month of Fitbit behavioral data to find the segment most likely to pay for Bellabeat's wellness membership.

RoleAnalysis & recommendations
ToolsPython · pandas · seaborn
ContextGoogle Data Analytics capstone
DateMay 2026
View on GitHub →
TL;DR

Only 21% of users consistently hit the 10,000-step target. The strongest acquisition opportunity is the 57% "almost-there" middle: users who track daily but miss their goals. The sharpest behavioral signal: sedentary time predicts poor sleep far more than activity predicts good sleep (−0.60 correlation), pointing membership toward movement micro-interventions, delivered when motivation dips on Sunday afternoons.

21%
Very Active
57%
Almost-there middle
22%
Sedentary

Activity-tier distribution. The 57% almost-there middle is where premium guidance creates the most value.

The business task

Bellabeat Membership is the company's recurring-revenue product. Hardware sells once, membership sells every month. The question I set out to answer: which user behaviors signal a willingness to pay for premium personalized guidance, so acquisition spend targets the right audience?

Findings

Engagement segments exist, and most users miss the 10K mark

Engagement segments by average daily steps

Users split into four activity tiers. Only one in five is Very Active; the rest sit below the 10,000-step target.

Implication

The two extremes are wrong targets: Very Active users are already engaged, Sedentary users aren't ready. The 57% in the middle is where personalized guidance creates the most value.

Sleep tracking adoption defies the obvious hypothesis

Sleep tracking adoption by segment

Sleep tracking was highest among Sedentary users and lowest among Lightly Active ones, the reverse of what I expected.

Implication

Sleep tracking is a separate engagement dimension. That points to two acquisition strategies: a "balanced wellness" pitch for multi-feature trackers, and a "starter engagement" pitch for the casual middle.

Sunday is the lowest-engagement day

Average steps by day of week

There's a 17% gap between Tuesday's peak and Sunday's trough, a weekend disengagement arc that resets on Monday.

Implication

Sunday afternoon is the highest-leverage delivery window: catch users when self-motivation is lowest and pivot them toward Monday's renewed intent.

Sedentary time predicts poor sleep more than activity predicts good sleep

Sedentary minutes vs sleep

Across 410 paired days, sedentary time correlated with sleep at −0.60, far stronger than steps (−0.19) or active minutes (−0.09).

Implication

Membership should emphasize movement micro-interventions (standing reminders, short walk prompts), not workout intensity. This fits Bellabeat's holistic wellness positioning.

What I'd recommend

1

Target the "almost-there" middle

Focus acquisition spend on the 57% in the Lightly and Fairly Active tiers: daily trackers who aren't hitting goals, exactly the audience premium guidance serves best.

2

Reframe membership around movement, not fitness

Lead with micro-interventions that reduce sedentary time, positioned as the path to better sleep. Aligned to Bellabeat's brand, not athletic-performance competitors.

3

Deliver premium content on Sunday afternoons

Concentrate motivational content and weekly planning into the week's lowest-motivation window. A defined day, time, and behavioral rationale, not just "engage more."

Honest caveats

Next step: with Bellabeat's own first-party data, I'd test whether this segmentation predicts actual conversion and retention, moving from directional insight to evidence-based strategy.