The Art of Analytics: Using Data to Drive Business Decisions

The art of data analytics is like cooking. You need the right ingredients, the perfect timing, and a pinch of creativity to make something truly remarkable. Except instead of using herbs and spices, you’re using data to create actions. And just like cooking, sometimes you have to experiment with different combinations of data to find the perfect recipe for success.

Photo by Mariusz Pierog on Unsplash

Meet Divvy, a ride-share company based in the windy city of Chicago. Their mission is to provide a convenient alternative for consumers to travel within the Chicago metro. To do this, they have both pedal and electrical bikes ready on demand for rent to consumers. Divvy’s subscription model is in the form of casual (one-time or day pass) payments or annual memberships. Rather than choosing business directions based off hunches, gut feelings, or emotions; you can use data analytics to forecast trends and predict probabilistic outcomes.

Divvy is a fair-weather business thus, they experience most of their annual revenue during warm summer months. Thus, they had introduced an annual membership to offset fallen revenue during the cold, windy and icy months. In the interest of converting more of their casual basis to annual pass memberships they approached, you, the data analyst.

As a data analyst, you, are given creative liberty to meet the aforementioned goal by employing trademark skills in the domains of data mining, processing, analysis, interpretation and visualization. You begin collecting the data between 01FEB2022–31JAN2023. Shortly after, you discover various correlations and inferences supported by both internal and external data, which include:

- Longer trips are more likely to be casual users (average duration of 12 minutes vs. 22 minutes).

- weather has a strong negative correlation for both client types on rides per day.

- Casual riders are more likely to use Divvy services on weekends rather than weekdays. (see graph below)

- About 59% of total bike rides by casual riders were via electric bikes, compared to about 49% for total bike rides by users with an annual membership.

After having analyzed the data and preparing various visualizations using software tools like Tableau and Visio, you, the data analyst, have proposed three data-driven actions for Divvy executives to consider. The first opportunity would be to implement a new annual pass tier catered to casual users. This proposed tier would use the benefits from the annual pass but restrict usage to weekends only. The next opportunity would be to market a promotion for casual users to experience the classic bike, which is the slightly preferred option of annual pass holders and the least favorable among first time users. The final recommendation is to expand electric bike holdings both in location and availability as it’s overall the most popular service Divvy users opt for.

Data-driven recommendations:

• Implement a new annual membership tier for weekend only usage at a lower price point.
• Offer an introductory promotion to casual users for first time classic bike rentals.
• Expand electric bike holding locations and availability.

About the data:

• NOAA (https://www.ncei.noaa.gov/cdo-web/datasets/LCD/stations/WBAN:94846/detail)
• Provided under this license: (https://nauticalcharts.noaa.gov/data/data-licensing.html)
• Divvy (https://divvy-tripdata.s3.amazonaws.com/index.html)
• Provided under this license: (https://ride.divvybikes.com/data-license-agreement)