/BH-PCMLAI-Module_5

This project contains the Module 5 assignment

Primary LanguageJupyter Notebook

BH-PCMLAI-Module_5

This project contains the Module 5 assignment and activities.

Overview

To answer the question: How would you determine whether a driver is likely to accept a coupon? We have explored the driver's profile of thuse who recived a 'Bar' and 'Coffe House' coupons and accepted it. We looked at drivers who had passengers other then kid in the car, are going to bar or a coffe house at least once a month, at different age groups, marital status and income.

summary of findings

Our driver profile is a person who is between the ages of 25 and 30 years old. They most likely visiting a bar or a coffe house, less then 3 times a month and not widowed. They are not working in farming, fishing, or forestry and respect the law.

In total our driver profile represent only 1.57 % of total drivers who accepted the bar coupon and 4.55 % of total drivers who accepted the Coffee House coupon.

The acceptance of the Bar coupon is less then half of the Coffee House coupon.

In both cases more female have rejected the coupons then man.

Most of those who accepted a coupon where either Single or Married.

Students and Unemployed people are the largest groups that accepted the coupon.

From those who accepted the 'Bar' coupon, the group with 'Some Highschool Education' dominated.

In additional exploration on the Bar coupon I asked

  • Is it likely that they go to more upscale restaurant, like 20-50?
  • Does the income change the profile?
  • Does gender have an impact if we ignore the income?

If we drop the income qulifications, then the number of coupon accepted by a diver increases segnificantly.

Reccomendations

Given the finding and additional observations my recomendations are to combine coupons, for example if the Coffee House, Restaurant (<20) and Carry Out & Take away coupons were combined they represent 79.4% of all coupons. 55.42% of these have been visited these type of businesses at least once a month.

Links

The link to the jupyter notebook used, supporting these findings can be found here