Identifying Trends and Relationships Within Smart Device Data
Summary
Bellabeat is an up and coming wellness company looking to promote various healthcare products. There aim is to improve the lives of woman across the world. In order to do that, they must be able to market their products effectively. In this case study I looked into data pertaining to health wearables. After exploring the data found, I was able to recommend marketing strategies for Bellabeats “Leaf” product, which is their man wearable. The project was done using “Excel” for data cleaning, “MYSQL” for data analysis, and “Tableau” for data visualization.
Intro
Who is Bellabeat? Bellabeat is wellness startup company. They primarily market several wearables, as well as other smart devices, to improve the health of woman. One of their products in particular, the “LEAF” bracelet, has been shown to be one of their top sellers. I wanted to explore any trends or relationships within smart device data to see how Bellabeat could continue to improve their market strategy for their LEAF bracelet.
Background Information
Smart device usage is becoming more apparent in our day and age, specifically when it comes to tracking one's health. Companies such as Bellabeats understand the importance of one tracking their health and the advantages of doing so. Statistics have found that about 30% of adults utilize some sort of smart device to track their health. This statistic shows that many people have already adopted the idea that wearables to help to better improve their health, however, it has not been accepted my most yet. Therefore, I wanted to look through past smart device usage data to see if I could uncover any trends or relationships that could better improve the market strategy for Bellabeats.
Business Objective
Identify any insights or trends within smart device usage data to help improve the marketing strategy for Bellabeat, specifically their “Leaf” product.
The goals I aimed to achieve through this analysis were to:
Identify trends within the data that align with Bellabeat’s products.
Suggest improvements, based on my analysis, to the marketing strategy for Bellabeat’s “Leaf” bracelet to ultimately improve revenue.
Preparation of The Data
Due to privacy reasons, I was not able to share the actual data from bellabeats, however the dataset used replicates it enough to mimic similar results.
Thirty eligible Fitbit users consented to the submission of personal tracker data, including minute-level output for physical activity, heart rate, and sleep monitoring. The dataset contains 30 days of user data that includes information about daily activity, steps, and heart rate that can be used to explore users’ habits. I decided to focus on two datasets. The two datasets were daily activity and sleep activity of the users.
The dataset was pulled from Kaggle.com
Some limitations to the data that I have found:
we are not sure whether the users are all women, mixed gender, or all males. With this being unclear the results may be affected.
This dataset is that it only contains data on 30 users. This is a small percentage of users and could possibly produce results that are different, had it been more recorded data from more people.
The dataset is from 2016. A dataset that is more relevant has the potential to produce different results from the ones found.
Data Processing & Cleaning
I conducted some simple data cleaning in excel in order to ensure that the data is consistent and prepped for analysis. I began my cleaning by removing any duplicates that were found from both datasets.
I then checked for any extra whitespaces, and finally filtered each column to check for any null values.
I noticed that the date column within the “sleepdata” dataset contains both the date and time, whereas the “activitydata” dataset has just the dates. Since the time in each cell within the “sleepdata” datset = "0:00", I decided to remove the times in order to make it consistent with the date column in the “activitydata” dataset.
Analysis
Once I completed my data cleaning, I uploaded each dataset into MySQL. This is the application I used to explore the data provided.
Queries used to help accomplish the business objective are found here:
Findings
1. Based on the bar chart, just over 35% of the user data recorded over 10,000 steps being taken. This number can play a role in the fact that those who wear a fitness watch are more likely to workout because they have something that is recording their fitness level that they can easily see. On average
2. When conducting my analysis, the data revealed to me that users who burned less calories throughout the day also slept less. On average, those who slept less than 6 hours burned around 2,131 calories vs 2,441 for those who slept between 6-8 hours. This can prove true because calories are burned even when one sleeps. The more an individual spends time sleeping, the more they are able to burn. This is due to the body repairing itself from the activity its been through. When an individuals spends a lot of time being active the body requires more time and need for recovery. Therefore, the longer one’s asleep the longer time their body has to repair.
Marketing Recommendation
Bellabeats could create an advertisement revealing how wearables could get users to become more active on a day to day basis. This being that since users are aware that they have a wearable on them tracking their daily activity, they will be more inclined to be more active throughout the day.
Bellabeats could also create an ad campaign for their “Leaf” watch revealing statistics that was found from the data how, on average, those who sleep more, burn more calories. Since their “Leaf” product has the ability to track both, this would be a great feature to market to their target audience.
Sources
Wearable Statistic:
Fitbit Dataset: