Proposal2

Based on the feedback update the proposal you wrote. Please follow the guidance based on the attached documents.

Please track changes in Word when editing your draft. You can go to Review – > Track Changes. You do not need to comment on where you added changes or why, just make the changes you think are most appropriate and track them.

Business Problem

The business problem related to the work is poor customer relationship management, which affects the business’s success. At the workplace, I experienced that our customer management strategies are not advanced and fail to satisfy customer expectations on time. This greatly affects the retention and returning rate of customers. This problem can be addressed through a data science solution. This proposal will determine effective data mining techniques that will address the problem and add value to the organization. The proposal will discuss the use of predictive modeling in solving the issue by using data from the organization.

Predictive Modeling

To address the problem, predictive modeling is helpful because by using existing customer status, the management can predict future outcomes and make changes to improve these outcomes. By analyzing existing customer relationship patterns, management can examine the organization’s likelihood of success or failure. But to utilize predictive modeling to solve CRM, the company should have a lot of customer interaction data or records. For example, records of 100 customers would help generalize customer experiences and design advanced and effective customer relationship management strategies.

Outcomes of Solution

Purpose of solving this problem is to help the company achieve long-term positive outcomes and develop a long-lasting positive relationship with customers. Positive customer relationships with the organization will add value by improving profitability and recognition. Customer retention and loyalty also bring huge advantages to the organization, which is what the solution wants to support and adds value to the organization.

An Integrated Model and Techniques of Data Mining

An integrated data mining model is helpful to address the problem, consisting of three types of data mining processes:

· Discovery

· Predictive Modeling

· Analysis

The data mining model’s discovery stage helps identify new customers and the best ways to interact with them. Data mining techniques such as segmentation and association will help identify customers and interact with them using the best techniques possible. The predictive modeling stage will help better understand past and present customer behavior. And the technique response modeling will help set new strategies and goals regarding CRM. The data mining model’s predictive modeling stage will also help attract and retain profitable customers for a long time. The analysis stage of the integrated model will conduct in-depth customer analysis and guide the organization on whether to retain customers for long or not. At the same time, techniques such as deviation detection and churn detection help determine deviation from the norm and help the organization make decisions.

The data required to train the model would be extracted from the organization. The organization required as much data as possible about customer relationships for predictive modeling. For predictive modeling, data scientists must take care of customer data, experiences, and satisfaction to help manage customer relationship management.

You should consider the feedback of your classmates when updating your proposal and there should be some degree of revision, updating, and improvement. Please track changes in Word when editing your draft. You can go to Review – > Track Changes. You do not need to comment on where you added changes or why, just make the changes you think are most appropriate and track them.

Feedback1

In your first section Business Problem, I think it would be great if you state the specific problem to the company, you would be presenting this proposal to. It seems like a general problem that any business might have. In the second section “Predictive Modeling” you said “to solve CRM” which is a confusing statement to me. Did you mean you will use the data from the CRM to help solve the problem? This section still feels very general and specific to the organization you are presenting the proposal to. The end of your proposal seems like a synopsis of data mining techniques and their advantages, but it does not define the attributes that you plan to use or a target attribute. At the end of the proposal, I am still asking myself exactly what type of business this proposal is meant for and what questions are you going to answer for them.
Thank you for letting me review your proposal.

Feedback2

Your Proposal was interesting, particularly on analyzing the current customer relationship patterns in the organization to improve or satisfy customer expectations on time. This can determine the firm’s likelihood of failure or success. However, there are a few suggestions to enhance your proposal based on CRISP-DM method.
Business Understanding
• Customer relationship patterns. What type of customer relationships does it involve? Does it entail value-add, community-based, emotional or transactional customer relationships?
• What types of customers’ records or responses will the organization consider for analyzing the feedback?
• Define your target variables. Your organization’s problem should state the target variables and focus your solution of the variables to come up with a conclusive answer.
Data Understanding/Data Preparation
• Your research on data was exciting because you identified 100 customer data to help in your predictive modeling.
• Attributes on past and present customer behavior

Feedback3

I think you have presented a very interesting data proposal. You clearly discussed the context of your business problem and its impact on your workplace regarding customer relationship management. I would, however, suggest listing what specific problem you are trying to solve in order to enhance your project understanding (i.e. “The problem at hand is an inadequate rate of satisfying customer expectations on time”). The proposal successfully outlines that the company will need to obtain/extract customer interaction data or records to execute the predictive modeling, indicating that you have a solid data understanding. One suggestion I have would be to choose a specific model/data mining technique type (i.e. decision tree, logistic regression, etc.) to use for the proposal and explain why you chose it over other options. This would be helpful in determining what the model would show, be used for, and how it will be evaluated in addition to clearly defining a target variable– which is currently a bit difficult to identify. Another suggestion I have would be to select five attributes that the model will examine to help explain the data that will be collected and the overall use of the model. This will help improve your data understanding and preparation. However, I did enjoy the outline of the integrated data mining model provided as it did give insight into what the model will identify and how it can be used to select the best customers and interaction techniques for the company. Overall, I think this data proposal has a strong framework with a solid explanation of how the model will function and be used but could benefit from a bit of clarification and expansion.