Strategies

Now that we have an accurate predictive model and baseline simulation we can determine which drivers should be adjusted, the distribution of values that will be needed and the strategies that will yield the target distribution.  We begin this effort by eliminating those drivers from consideration that cannot be properly adjusted based on cost, quality or speed constraints.

For the remaining drivers, we develop the strategies that can be used to make an adjustment, set the distribution of values and run a target simulation. The simulation is optimized by adjusting the drivers so that the intended outcome will occur.  It is likely that there will be multiple combinations of distributions that result in the intended outcome.  The combination that has the best chance of succeeding, given the feasibility of the strategies that are needed, is the one that is chosen.

Our strategies will have to ensure a certain level of cost, quality or speed of products and services.  In addition, they will target the consumers that are most likely to make use of our products and services given these constraints. 

The example that we use to explain the Weincor business intelligence method is as follows (1)

Students that had been offered college admission are surveyed and asked to rate the importance of various factors that may have affected their decision to accept or reject the offer using this scale; 1 - Not important, 2 - Somewhat important, 3 - Important, 4 - Very important and 5 - Critical to the decision.  Prior analysis had identified the drivers that are statistically significant.  Unfortunately some of these drivers cannot be properly adjusted based on cost, quality or speed constraints.  Each of the drivers have a prefix (short name) and a description as follows

a10 - Great parties

a22 - Close to home

a24 - College offers spring break (for course credit) in cool countries

a41 - Large college

a56 - Hearing from the college president by telephone

a70 - College will accept all/most of credit earned elsewhere

A decision to accept an offer of admission based on great parties, close to home or large college is very subjective and it would be difficult for the college to build this type of brand recognition.  It would also be difficult to find time for the college president to reach out to each student that is offered admission.  The college's quality of instruction exceeds many of it's peers and it would have to lower it's standards if it were to accept all/most of credit earned elsewhere.  However these remaining drivers can be affected by appropriate strategies.

a13 - Attending an event (e.g. theatrical, sporting, homecoming...etc.) with a faculty member

a30 - Housing quality

a38 - Friends recommendations

a54 - Meeting a faculty member in their office

a63 - Meeting a graduate of the college while shopping, working, going to a concert

A strategy is developed to provide opportunities to attend an event with a faculty member and to meet them in their office.  A strategy is developed to assess the quality of college housing and make any necessary improvements.  A strategy is developed to make use of social networking to provide easier access to recommendations from friends and opportunities to connect with alumni.  A strategy is also developed that first offers admission to students that are most likely to accept an offer of admission, if all else is equal.

A target simulation is created using the optimal distribution of values and the outcome shows a 5% decrease in the rate of admission offers that were rejected. 

The distribution of values for all of these drivers are survey responses of either 4 - Very important or 5 - Critical to their decision. Obviously these distributions have been set for the purpose of illustration.

The strategies that improve access to faculty, quality of housing and use of social networking will need to be completed before prospects have been identified.  At that time prospects will be surveyed and the data that is collected will be used to predict how likely they are to accept an offer of admission.  If the distribution of values and predicted outcomes do not match the target simulation, changes will be made to the colleges marketing efforts.  An applicant's likelihood of accepting an offer of admission will be considered when evaluating their qualifications.

(1) The data that was used to perform the target simulation was included in the book Dr. Green's New Hyper-Linked Introduction to IBM SPSS Modeler, Dr. John B. Green Jr. PHD, Left Brain Books, LLC.  The successful use of business intelligence depends on the proper identification and collection of data.  Dr. Green provides a valuable data resource to illustrate our method.