Project 2: Machine LearningApplications
Project Brief
SCENARIO:
Your manager, Andrew Frey, has deployed your team to assist a client with their efforts
in improving customer satisfaction by looking at underlying data. The client’s
expectation is that the team would help in streamlining data & analytics activities so
customer satisfaction can be improved with lower cost.
LiveWell is a healthcare managing company that is focused on overall wellness of their
customers. Every month the company analyzes customer satisfaction through net
promotor score (NPS). The customers are asked to rank their satisfaction level on a scale
from 1 to 10. On this scale higher score (9 or 10) is considered as an indication of happy
customer. Anything less than 9 is considered as an indication of unhappiness.
Currently, every month, an associate looks at the survey results and analyzes the results
of 10,000 unhappy customers with the help of several variables (such as issue resolution
time, answering calls, claims processing etc.) affecting customer satisfaction. Based on
this analysis drivers (variables) causing unhappiness are identified. Using these
results, strategies are designed with the hope that they will eventually enhance
customer satisfaction. This entire activity takes 20 hours associates’ time. After this, the
manager will also have to review these results to decide on deploying these
strategies. The manger review will take 5 hours. Assume a manager bills at twice the
hourly rate as an associate. The company thinks that current analysis is not very robust
and it is only 50% accurate, and it also does not consider data quality aspects. The
company also thinks the accuracy of the analysis is affected by poor data quality by 5%.
You’ve already developed a model that can be used to determine root cause drivers and
it can also be used to predict NPS based on the different drivers. Further, your model is
60% accurate. In addition, your model also has built-in data quality feature that will
ensure the data used for model build is of high quality and fit for the purpose. Running
this model takes 8 hours of an associate’s time for set-up , data pre-processing and
analysis. Because of the higher complexity of the this modeling approach, the manager
must spend a total of 3 hours reviewing the associate’s work.
You must make the decision on whether or not to use your model for this task and then
defend that decision to your manager. You want to focus on three considerations when
making your recommendation:
1. Minimizing the standardized hours required for this entire process,
2. Highlighting importance of data quality for analytics and
3. Providing better results through this model to enhance confidence in decisions.
KEY TASKS Make a decision as to whether or not you would recommend the implementation of your
model. Then write a report to Andrew explaining your decision. Be sure to include the
following:
1. Your final recommendation should focus on whether to implement the model or
continue with the current approach. You should include a justification of why you
chose your proposed solution over the alternative, including key measures of the
cost and performance of each approach.
2. An explanation about importance of having high-quality data.
3. If your team wants to further improve the accuracy of analytics, what types of
models/analytics would you recommend for this application?
4. Any other considerations you feel are relevant to this decision, or additional
information that would be helpful to have.
5. The quantitative analysis you performed to support your recommendation,
including an estimate of the standardized hours required under each approach,
reducing cost due to poor data quality and any model accuracy calculations that
were relevant to your decision.
6. Include all your references at the end of your document.
Keep in mind the following:
•
•
•
•
The executives reading the report are non-technical people and will not
understand jargon or technical speak. They have asked that you provide your
calculations as a reference, but the body of the report should summarize the
key results of your analysis rather than diving into details.
The executives will not likely read your report more than once and will be put
off by anything overly complicated or unfocused on their questions.
While you need to provide the information requested, it’s up to you to
determine how broad or specific the contents are and what content is
included.
Effective use of this project’s required materials is critical to your success.
WHAT TO SUBMIT:
Your report to Andrew explaining your decision regarding the implementation of your
model.
Please note: Since this report focuses on conciseness and relevancy of information,
please limit your report length to 8-10 pages.
Order an Essay Now & Get These Features For Free:
Turnitin Report
Formatting
Title Page
Citation
Outline
