Scribd
Upload
43 views17 pages

Operations and Metric Analytics - Case Study

Uploaded by

nitant
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
43 views17 pages

Operations and Metric Analytics - Case Study

Uploaded by

nitant
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 17

Operations and Metric

Analytics - Case Study

By - Nitant Sharma
Project Description - Case Study 1

-> In this project, we are going to help Operations Analysis team in


fetching the required information from the Job Data table.

-> With the help of operations analysis we can find the areas which can be
improved and can predict the overall growth or decline of a company’s
fortune.
Approach

-> First step was creating a database where the dataset would be stored.

-> Next step was importing dataset present in the form of csv to MySQL
Workbench.

-> The structure of the data which is to be stored inside the table job_data
was mapped with correct data type to avoid data loss while transfer.

-> Finally query operations were tested to perform analysis.


Tech Stack Used

1) MySQL Workbench 8.0 Community Edition


-> It allows us to design,create,run queries using a graphical user interface.
2) MySQL
-> It is as structured query language which is used to communicate with
database.
3) Microsoft Excel
Insights

Number of jobs reviewed per hour per day for November 2020
Select
throughput.total_jobs_per_day,
throughput.hrs_spent
FROM(
SELECT ds,
COUNT(job_id) as total_jobs_per_day,
sum(time_spent)/3600 as hrs_spent
FROM job_data
WHERE ds>='2020-11-01' and ds<='2020-11-30'
group by 1
)throughput;
Calculate 7 day rolling average of throughput

SELECT
throughput.ds,
throughput.total_jobs_per_day as Num_of_jobs,
AVG(throughput.total_jobs_per_day) OVER (ORDER BY throughput.ds ROWS BETWEEN 6 PRECEDING
AND CURRENT ROW) AS Seven_day_moving_avg
FROM(
SELECT ds,
COUNT(job_id) as total_jobs_per_day,
sum(time_spent) as hrs_spent
FROM job_data
WHERE ds>='2020-11-01' and ds<='2020-11-30'
group by 1
)throughput;
The percentage share of each language in the
last 30 days
Select
sub.ds,
sub.`language`,
sub.count_of_language AS Count_of_languages,
(sub.count_of_language) * 100.0 /
sum(sub.count_of_language) over(ROWS BETWEEN
UNBOUNDED PRECEDING AND UNBOUNDED
FOLLOWING) AS Percentage,
sum(sub.count_of_language) over(ROWS BETWEEN
UNBOUNDED PRECEDING AND UNBOUNDED
FOLLOWING) As Total_languages
FROM(
SELECT ds,
`language`,
COUNT(job_id) as count_of_language
FROM job_data
WHERE ds>='2020-11-01' and ds<='2020-11-30'
group by `language`
)sub;
Display duplicates from the table

Select *
From
(Select *,
ROW_NUMBER() OVER(PARTITION BY job_id ORDER BY job_id) AS Row_No
From job_data) dup_sub
Where dup_sub.ROW_No > 1;
Case Study 2 - Investigating metric spike

-> In this project we will investigate the dataset further for answering the
questions asked by the team by giving them insights of the data.

-> Investigating metric spike is an important part of operation analytics as being a


Data Analyst you must be able to understand or make other teams understand
questions like- Why is there a dip in daily engagement? Why have sales taken a dip?
Etc. Questions like these must be answered daily.
Insights
Calculate the weekly user engagement

SELECT EXTRACT(WEEK From


occurred_at) AS 'Week',
COUNT(DISTINCT user_id) AS
No_of_active_users
FROM events
WHERE event_type = 'engagement'
AND event_name = 'login'
Group by 1;
Calculate the user growth for product

Select Distinct user_id,


device as Device_Name,
Extract(Month from occurred_at) as
Month_No,
count(device) over(partition by user_id) as
User_Device_Interest_Growth
From events
Order By Device_Name,Month_No;
Calculate the weekly retention of users-sign
up cohort

Select distinct user_id,


count(sub.Week_No)
over(partition by user_id) as
Weekly_Retention_Count
from(
select distinct user_id,
Extract(Week from
occurred_at) as Week_No
from events
) as sub;
Calculate the weekly engagement per device

SELECT user_id,
device as Device_Name,
WEEK(occurred_at) as
Weekly_Engagements,
COUNT(Case when event_type =
'engagement' then 1 ELSE NULL
END) as No_Of_Engagements
FROM events
GROUP BY
user_id,Device_Name,Weekly_Enga
gements
ORDER BY Device_Name;
Calculate the email engagement metrics

Select Distinct user_id,


Action,
Count(action)
over(partition by action)
as Email_Engagements
From
email_events
Order By user_id;
Result
-> The jobs reviewed per day were not more than 2.

-> The 7 day rolling average lies between 1 and 1.3.

-> The Persian language had a maximum share of 37.5 among all languages.

-> The dataset contains two duplicate rows which might affect the result of data analysis.

-> From the weekly user engagement report it was known that user were least active on Week
No: 35.

-> Monthly users growth for product increase from Month No: 5 to 6.

-> Weekly users retention report will be helpful in understanding how many users are still active
on website after signing up.

-> We found out how many times users interacted with the website for the product in a week
using Weekly Engagement per device.

-> Email metrics report will help in knowing the action of the users with email i.e
sent_weekly_digest, email_open and email_clickthrough.
PPT Link
https://docs.google.com/presentation/d/1ED11eNUNn1rHUejVRAKBPNeM
uOVByrwBhTwwy2pDTdU/edit?usp=sharing

SQL Code Link


https://drive.google.com/file/d/1pNIBfwZKHPxCaEAZ6nhQgV5sXRpsUuIM
/view?usp=sharing

https://drive.google.com/file/d/1e659NUCNzyRQET3ZI3QDXC46v352aYck/
view?usp=sharing

Dataset Link
SQL Project-1 Table - Google Sheets

SQL Case Study-2 Data Set - Google Drive


THANK YOU

You might also like