Scribd
Upload
678 views

Data Analyst Case Study Example

Most popular days for ridesharing were Friday and Saturday. The peak hours were 6-7pm on weekdays and 10pm on Saturdays, indicating increased leisure activity on weekends. From March to June 2016, there were 184k rides by 844 drivers, averaging 219 rides per driver. Total revenue was $5.4 million with $1.08 million in profits. On average, drivers stayed 1.62 months and were worth $10,380 in revenue. The churn rate was low at -3.63% on average. Most drivers either outperformed or underperformed averages, with underperformers more likely to churn.

Uploaded by

mg2001
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
678 views

Data Analyst Case Study Example

Most popular days for ridesharing were Friday and Saturday. The peak hours were 6-7pm on weekdays and 10pm on Saturdays, indicating increased leisure activity on weekends. From March to June 2016, there were 184k rides by 844 drivers, averaging 219 rides per driver. Total revenue was $5.4 million with $1.08 million in profits. On average, drivers stayed 1.62 months and were worth $10,380 in revenue. The churn rate was low at -3.63% on average. Most drivers either outperformed or underperformed averages, with underperformers more likely to churn.

Uploaded by

mg2001
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 10

Case study

presentation –
Ridesharing
Analytics
peak
peak

Most popular
days peak

Based on the data, from late


March to June 2016, it
appears ride orders peaked
similarly across months:
Friday, followed by Saturday.
Most popular
hours - Highest number of rides per day occurred at 18.00 and 19:00

weekdays
Rides on 18.00 and 19:00 PM had
the highest occurrences per day.

When broken down based on


weekdays, it appears that Friday
and Saturday had the highest
number of rides most popular hour
in a weekday is Saturday 22:00,
followed by Friday 19:00. This
indicates ride frequency may
increase when people want to have
more leisure time.

Most popular hour: 22


on saturday
Number of
Rides per
month
During March-June 2016, there
were a total of 184K rides served by
844 unique drivers. Within this
period, each driver delivered 219
ride.

During the latest month there was


67K rides served by 679 riders.

With an average of 46K rides per


month, May and June had an above
than average within the time period.
Revenue
The latest month showed a revenue of
$1.930 million with a $386K margin
profit.

From March to June 2016, the total


revenue was $5.4 million. As the
company takes 20% pay cut, the total
margin profit during this period was
$1.080 million.

When broken down based on months,


the highest revenue occurred in May
2016, as the highest number of rides
occurred in the same month.
Average
Lifetime
Value
To calculate Average lifetime value per
driver we need to calculate the
following metrics: average revenue per
transaction, average lifespan, and the
total ride per unique driver

Based on the calculation of those


metrics, the Average lifetime value of
the drivers is $10.38K. It means
based on the existing data, we could
expect on average, each driver
contributed a $10.38K revenue during
the length of their partnership, which
translates as $2.08K gross profit.
Churn rate
Due to limitation of data, churn rate
is calculated on day-to-day basis. On
monthly average, there was a
negative churn rate of -3.63%,
which indicates that there are more
active drivers compared to drivers
who churned.

There was a spike of churn rate on


27th June 2016, however it appears
to be a result of incomplete data as
the latest data of the said date only
lasted up to 00:50.
Cohort
Table
Based on the three months
data, 69.79% of the drivers
kept working with the company
after three months.

On average, we could expect


drivers to stay working with the
company for 1.62 months.
Either you outperformed and stay

Driver
segmentation
On average, the driver made 219 rides and a
revenue of 6.4K. This visual shows that there is
a linear relation between the number of rides
and revenue.

Majority of the drivers fell between Quadrant I


(41%) or Quadrant III (52%).

This indicate most of the drivers either


underperformed or outperformed the average,
and only a very few drivers (2.24%) fell into the
lucky Quadrant II where they had less than
average rides yet had above than average
revenue.

Based on this segmentation, drivers in


Quadrant III had the highest churn in the latest
month. Which indicated drivers who
underperformed may tend to churn.
Or underperformed and churned
Conclusion
On daily basis, the highest hours appeared to be 18:00 &19:00, while Saturday 22:00 the highest number of rides during
weekdays, which indicate people are willing to pay for convenience during weekends. This information can be used to adjust
the pricing as the traffic likely to surge during those “prime time”.

The total revenue was 5.4 million, with 20% margin profit, the company made $1.08 million across 184K rides, which were
served by 844 unique drivers. On average, each driver delivered 219 ride. The average revenue per transaction was 29. The
average driver stayed with the company was 1.62. The lifetime value of each driver is $10.38K, and each driver brought
approximately $2.08K of margin profit.

During the period, churn rate was not bad: on daily average, there was a negative churn rate of -3.08%. Negative here
indicates there are more active drivers than those who churned. Based on cohort table, 67.9% stayed with the company
after three months.

When segmented based on their average ride and average revenue, most drivers fell either underperformed or
outperformed the average. Which indicates there’s an even distribution between underperformers and outperformers. This
affects the churn rate as well. It appears drivers who underperformed on rides and revenue have a higher churn rate. This
information can be used to determine whether drivers who underperform could be incentivized so they will stay longer.

You might also like