Raj HamalData Analyst
Back to All Projects
Applied AI / CapstoneIn Progress2026

Bellabeat Smart Device Usage Analysis

Google Data Analytics Capstone – Smart Device Health Data Exploration (In Progress)

Project StatusIn Progress
CategoryApplied AI / Capstone
Primary ToolsGoogle BigQuery, SQL
AuthorRaj Hamal

Business Problem & Objectives

Bellabeat, a high-tech manufacturer of health-focused smart products for women, wants to analyze smart device usage data to uncover how consumers use non-Bellabeat smart devices (Fitbit). Findings will guide executive marketing strategy for the Bellabeat Leaf wellness tracker.

Dataset Overview

Fitbit Fitness Tracker Data (Public Domain via Kaggle/Mobius), containing personal fitness tracker responses from 33 eligible Fitbit users including minute-level activity, heart rate, and sleep monitoring data.

Data Preparation & Cleaning Steps

  • Loaded raw Fitbit CSV datasets (dailyActivity, sleepDay, weightLog) into Google BigQuery environment.
  • Queried schema definitions and checked for duplicate user ID entries across 33 distinct participants.
  • Standardized date timestamps into YYYY-MM-DD formats using SQL CAST functions.
  • Calculated daily active time categories: VeryActiveMinutes, FairlyActiveMinutes, LightlyActiveMinutes, and SedentaryMinutes.

Analytical Methodology & Core Insights

Methodology Framework

  • •BigQuery SQL Aggregations: Grouped daily step counts, calorie burn, and sleep duration per user.
  • •Activity Segmentation: Classified users into activity tiers (Sedentary < 5k steps, Low < 7.5k, Somewhat Active < 10k, Active > 10k).
  • •Correlation Analysis: Evaluated relationship between daily active steps and sleep quality indicators.

Analysis Highlights

  • •Sedentary Dominance: Participants spent an average of 991 minutes (~16.5 hours) per day in sedentary state.
  • •Step Count Clustering: Average daily step count was ~7,637 steps, slightly below the recommended 10,000-step daily target.
  • •Sleep & Activity Link: Initial SQL queries indicate a positive correlation between active afternoon movement and sleep efficiency score.

Key Findings

  • In-Progress Exploration: Ongoing BigQuery analysis focused on hourly activity spikes during 5 PM – 7 PM commute windows.
  • User Engagement Drop-off: Tracking compliance decreases significantly after 20 consecutive days of logging.

Business Recommendations

  • Smart Notification Triggers: Send gentle sedentary alerts on the Bellabeat app during mid-afternoon idle hours.
  • Gamified Habit Loops: Introduce weekly streak rewards to maintain user logging compliance past the 20-day drop-off threshold.