联系方式

  • QQ:99515681
  • 邮箱:99515681@qq.com
  • 工作时间:8:00-21:00
  • 微信:codinghelp

您当前位置:首页 >> Java编程Java编程

日期:2020-10-19 10:59

Study Case for Claims Analytics Candidates

The Co-operators - Claims Portfolio Analytics

October 2020

1 Instructions

This study case is illustrative of a typical business problem that the Claims Portfolio Analytics team tackles at the Co-operators. We are

interested in observing how you approach the problem, develop a solution and provide insights and recommendations.

Candidates are allowed 5 days to work on the study case in preparation for the interview;

During the interview, you will be given 10 minutes to tell us about your observations and formulate recommendations;

We expect a Powerpoint-style presentation with supporting charts, graphs and reports;

At the end of the presentation, we will be interested to understand the work that you did leading to the presentation.

No specific tool is mandated, so feel free to use the tools that you are familiar with and appropriate for the study case.

The presentation and ensuing discussion will be evaluated according to:

Strenght and relevance of insights and recommendation;

Adequacy of the underlying analytic work (right solution for the right problem);

Clear and effective communication.

We acknowledge the time and other constraints that candidates may have in preparing for the interview. It is not our expectation that candidates

invest more than a few hours, and the evaluation will be done accordingly. We invite you to spend your time wisely, for instance by focusing on

one key result/recommendation rather than many.

If you have any clarification question about what is expected in this study case,  before Tuesday, Oct 20th, 5:00 pm EST. All candidates will see

responses to all clarification questions that will be received.

Good luck!

2 Business Problem Statement

Imagine that you work for an insurance company wanting to better understand the satisfaction of their clients, as it relates to their recent claim

experience.

The Claims Management team has requested that you explain the results of the company for the past several months and what we can learn

from surveys sent to claimants. More specifically, they are asking your help in identifying unusual trends and potential opportunities for

improvement.

3 Data

The claim.sat dataset from the claim_sat.csv file contains the response received from claimants who were surveyed shortly after their claim

has closed, along with some relevant contextual information for a fictional insurer.

head(claim.sat, n = 3)

## rid Claim.Number HVI Date.of.loss Product Sales.region

## 1 4497 1653548 2 2019-08-02T00:00:00Z PROPERTY WESTERN REGION

## 2 7006 1656454 1 2019-03-07T00:00:00Z PROPERTY CENTRAL ONTARIO REGION

## 3 757 1636951 1 2019-07-08T00:00:00Z AUTO NEWOR REGION

## District Writing.company Days.to.close Base.direct.payment

## 1 YELLOWHEAD NORTH CGIC 13 6968.42

## 2 YORK/DURHAM CGIC 24 7761.58

## 3 WESTERN ONTARIO SHORES CGIC 23 3145.59

## Date.survey.input Care.and.empathy Knowledgeable.service Timely.updates

## 1 2019-08-31 3 2 3

## 2 2019-09-17 3 3 1

## 3 2019-08-08 2 2 3

## Timely.settlement Easy.claims.process Vendor.satisfaction Likely.to.recommend

## 1 2 4 0 2

## 2 3 8 0 8

## 3 1 1 1 1

Respondents are answering questions (on a 1-unfavorable to 5-favorable scale) such as whether the claim representative(s)

demonstrated care and empathy throughout the process ( Care.and.empathy );

were knowledgeable ( Knowledgeable.service ) ;

provided timely updates ( Timely.updates );

ultimately led the file to resolution in a timely manner ( Timely.settlement ).

Claimants are also asked if the claim process was made easy for them ( Easy.claims.process ) and if they were satisfied with the vendor

( Vendor.satisfaction ) (when at least one was involved, as indicated by a response greater than 0). The higher the score is, the more positive

the experience.

Finally, they are asked if based on their overall experience, they would recommend the insurer and products to their family & friends

( Likely.to.recommend ). Likelihood to recommend is associated with a well-known metric called Net Promoter Score

(https://en.wikipedia.org/wiki/Net_Promoter). It is popular across many industries to measure client loyalty.

The contextual information includes several characteristics about the product and the claim.


版权所有:编程辅导网 2021 All Rights Reserved 联系方式:QQ:99515681 微信:codinghelp 电子信箱:99515681@qq.com
免责声明:本站部分内容从网络整理而来,只供参考!如有版权问题可联系本站删除。 站长地图

python代写
微信客服:codinghelp