/cooker_sale_analysis

Primary LanguageJupyter Notebook

Cooker Sale Data Analysis

I have added the cleaned datasets in xlsx files to the repository. The context this project is based on can be found below...

Part I

Context

BURN’s business uses a “razor and blades” model, where it earns the majority of its profit by selling BURN Fuel. BURN Fuel can only be used in a BURN Cooker, and BURN Cookers can only operate with BURN Fuel. Therefore, BURN must first sell customers its Cooker and then ensure they have a positive experience so that they adopt BURN as their primary cooking solution and continue to buy Fuel regularly. The BURN Cooker is sold through three channels:

  1. Direct Sales: this channel consists of full-time BURN employees called Customer Sales Reps (CSRs) who sell cookers door-to-door in an assigned territory. They are paid a commission for each cooker sold plus a fixed monthly salary. They receive in-depth product and sales training and have access to demonstration cookers to use during their sales pitch. In total, BURN has 50 CSRs.
  2. Agents: this channel consists of shopkeepers who own and operate small retail outlets where BURNpoint fuel dispensers are installed. They are paid a commission for each cooker sold as well as commission for fuel sold. They receive basic product training. In total, BURN has 500 Agents.
  3. Referrals: this channel consists of existing BURN customers who sell cookers as a side hustle. They are paid a commission for each cooker sold. They receive no product training. In total, BURN has 10,000 customers. Exercise The first attached data set contains three tables:
  4. Sellers: data with details on all individuals active across BURN’s three sales channels.
  5. Cooker Sales: data with details on each cooker sold in January 2019.
  6. Fuel Sales: data with details on all fuel sold from January - March 2019. BURN wants your help to better understand the performance of these three sales channels: • What is the sales productivity for each channel? o Which channel is the most productive? o Is the most productive channel also the most efficient channel? Why or why not? o Whatopportunitiesmightexisttoincreasesalesproductivity? o What additional data would you want to better answer the above questions? • What is the average monthly fuel consumption per cooker for each channel? o Which channel produces the most satisfied customers? o Why might fuel consumption differ by sales channel? o Whatopportunitiesmightexisttoincreasefuelconsumption? o What additional data would you want to better answer the above questions?

Part II

Context

Given the context for BURN’s business in Part I, BURN must endeavor to retain customers and keep them active! Monthly Active Users is a core metric for the organization, and you will be responsible for providing insights around this and other related retention metrics. Monthly Active Rate Active Rate defines the proportion of fuel users who have filled their canisters out of the entire fuel user population within a given measurement period. Methodology: Active Rate = [number of fuel users who have filled canisters] / [number of fuel users] • Total fuel users who have refilled canisters: total number of fuel users who have refilled canisters within a given measurement period o Inclusioncriteria: ▪ Users who have had a fuel canister greater than the measurement period i.e. exclude ‘New Users’ • Total fuel users: total number of fuel users within a given measurement period o Inclusioncriteria: ▪ Users with a customer tenure (Days since fulfillment) greater than the measurement period i.e. exclude ‘New Users’

The second attached data set contains three tables:

  1. Customer Metrics: Rolling summary data with one row of data per month of fuel consumption. This is data for seven months.
  2. User: data with details of each BURN registered user
  3. Neighborhoods: data with details on all BURN Neighborhoods. BURN wants your help to better understand customer behavior based on Monthly Active Rate:
  4. What is the Monthly Active Rate and what can you say about the trend?
  5. Which neighborhoods / networks have higher retention?
  6. Which neighborhoods have the lowest retention?
  7. Is there anything to be said about retention and the number of BURN Points in the various neighborhoods?
  8. BURN defines customer cohorts based on the Quarter and Year of Fulfillment e.g Q1-2020, Q2- 2020 etc. Is there anything to be said about retention and Cohorts?
  9. What other hypothesis do you think could be tested to understand retention?
  10. What additional data would you want to better answer the above questions?