Case Study : Marketing Campaigns For Customers

1679 Words7 Pages
Chen Liu Professor Joe Dery Homework 3 Nov 4th, 2015 RFM Analysis A. Executive Summary Joe’s Shopping World schedules to launch marketing campaigns for customers. To maximize return of investment of each dollar on campaigns, Joe’s Shopping World is looking to profile the customer segmentation to test incremental response from the marketing campaigns. The dataset contained 225,533 rows of customer transaction records over the last two years. The following elements have been recorded in the dataset: household ID, day (1-711), sales value ($) and total discount ($). I adopted the RFM analysis in this report to analyze the dataset and score each customer based on recency-the number of days since the last purchase, frequency-the number of…show more content…
• Frequency is the total number of days per customer visited in the last two years and is assigned the scores from 0 to 4. • Monetary is the total spending in the store over the last 711 days and is also assigned the scores from 0 to 4. 1’st, using SAS coding to calculate recency. 2’nd, grouping R, F and M records by household ID. Then there will be a new dataset that each customer have a row of record including recency (R_Cal), frequency (F_Cal) and monetary (M_Cal), as shown in figure-1. At the same time, I present a part of the original dataset to compare with figure-2. Figure-1: Query Dataset Figure 2: Part of Original Dataset 3’rd, ranking the data with separate customers into 5 groups based on R_Cal and assign R score to each customer. Of course, we need to check the reverse ranking radio-button to make sure the minimum value of recency has the largest value when we rank the customers. Because we use “recency=712-day” to calculate the R score. Part of the result is shown in figure-3. Figure-3: R Rank 4’th, ranking F and M at the same time. Just repeat step 3, but do not forget to uncheck the reverse ranking radio-button. Figure-4 shows part of the resulted dataset. Figure-4: F and M Rank 5’th, summarizing the result R Rank. We place R_Cal as the analysis variable and R as the classification variable. In figure-5, we can see the summary statistics with R_Cal. And the range of
Open Document