ANLY FPX 5510 Assessment 3

Comparison of Analytic Methods 

A large financial services company’s marketing department recently had its budget reduced and came to its analytics department for its assistance in the best utilization of their existing budget. The analytics department was requested to examine the communication plans of the company with different customer groups. The main objective was to avoid contact with potential customers who were unlikely to open an account with the company and save on wasted marketing expenses.

ANLY FPX 5510 Assessment 3:Compare 

The analytics team proposed two alternatives. One was applying cluster analysis to establish whether customers were being marketed through several channels, such as telephone, email, Facebook, Twitter, Instagram, Pinterest, Yahoo, Google search, and US Postal Service mail. The analysis sought to segment the customers with common behaviors to reduce the number of communication channels being used by the company. The results showed that customers preferred some types of communication. Based on these results, the team proposed maintaining advertisements via phone, mail, and Google and deactivating other types of communication. The second recommendation was through decision tree analysis. Just like cluster analysis, this also explored if customers were being advertised via the same media mentioned above. However, the decision tree analysis also encompassed demographic information and numbers of the total deposits customers had made in the previous year. The concept was to identify characteristics that would allow the marketing department to target high-deposit customers. From the analysis, the group recommended retaining Instagram advertising for clients under the age of 43. For clients aged 56 to 80, it was advised to discontinue advertising since this group had a lower percentage of high deposit values in the past year.

Appropriateness of Analytic Methods 

Cluster analysis is a set of techniques for partitioning objects or cases into clusters.

This method is commonly applied in marketing to establish groups of consumers with similar properties. The process typically involves some steps, including the specification of the problem, the selection of a measure of distance, the selection of a clustering algorithm, the identification of the number of clusters, acquaintance with cluster profiles, and evaluation of cluster validity. In the present business case, cluster analysis was an appropriate technique, as it effectively grouped customers based on their behavior, which in turn informed the company’s marketing and advertising policies (Statistics Solutions, 2019). On the other hand, decision tree analysis displays various alternative solutions to an issue graphically and is widely used in organizational decision-making. When faced with a choice, management is able to predict various solutions, each of which can be graphically depicted as part of a decision tree in order to simplify the decision-making process (Mulder, 2017). In this business case, decision tree analysis was a suitable choice for data analysis based on certain criteria. The method was utilized in order to make it possible to target customers of higher deposit values, which ultimately resulted in concentrated advertising campaigns.

Recommendations 

The recommendation for how the business problem can be addressed is the method of cluster analysis.

This method is based on identifying the most effective channels of communication instead of completely excluding some customer segments.

By analyzing the effectiveness of different advertising mediums, such as telephone, email, Facebook, Twitter, Instagram, Pinterest, Yahoo, Google search, and US Postal Service mail, cluster analysis helps determine which means are least expensive for marketing to customers.

Decision tree analysis, while effective, was concerned mostly with customer demographics and recommended not sending adverts to some age groups. Cluster analysis also has some advantages, such as being less costly. It is often less expensive to observe clusters in a population compared to the random selection of units distributed over a large area. Apart from that, cluster analysis also allows the company to take larger samples, which means it becomes simpler to obtain data in different areas (Money Matters, 2019).

Implementation 

It is not a one-off process to integrate analytics into an organization but rather an evolution of data-driven business processes.

With innovations increasingly specialized for the purposes of customers, products, and business goals, analytics-driven solutions will be an integral part of decision-making. To implement the results of cluster analysis, it is advisable to adopt a centralized approach. A centralized approach would be to place all the concerned assets, such as software, hardware, expert talent, metadata, business relationships, and project management skill, under one umbrella.

A centralized method has benefits like knowledge sharing, one implementation, better support, visibility, and clear channels of communication. It is cost-saving as it shares software and hardware on multiple projects. The first step in achieving this would be installing the cluster analysis data and process. The next step would be training the staff in the necessary methods and data handling. With this as the foundation, a project development methodology would be established before actual projects are undertaken (Sheikh, 2013). Monitoring and maintaining the cluster analysis will be essential to ensure that the company continues to utilize its findings in the correct way.