Marketing and Customer Relationship Management 2 Business Context for DM Although the technical aspects of DM are interesting and exciting (at least to geeks!), they must be utilized in a business context to be of value. Business topics addressed in this chapter are roughly in ascending order of complexity of the customer relationship, starting: Communication with prospects (little knowledge of them) On-going customer relationships involving multiple: Products Communication channels/methods Increasingly individualized interactions 3 Prospecting Prospect Noun someone/something with possibilities Verb to explore > 6B people worldwide Relatively few are prospects for a company Exclusion based on geography, age, ability to pay, need for product/service, etc. Data mining can help in prospecting: Identifying good prospects Choosing appropriate communication channels Picking suitable messages 4 Data Mining & Advertising Who fits the profile for this nationwide publication? Reader- ship YES Score NO Score Mike Nancy Mike Score Nancy Score BS or > 58% 0.58 0.42 Yes No 0.58 0.42 Prof/Exec 46% 0.46 0.54 Yes No 0.46 0.54 $ > $75k 21% 0.21 0.79 Yes No 0.21 0.79 $ > $100k 7% 0.07 0.93 No No 0.93 0.93 Total 2.18 2.68 5 Data Mining & Advertising Butthat might be a bit nave; compare readership to US population, then score Mike and Nancy Mikes score: 8.42 (2.86 + 2.40 + 2.21 + 0.95) Nancys score: 3.02 (0.53 + 0.67 + 0.87 + 0.95) Reader- ship YES US Pop Index Reader- ship NO US Pop Index BS or > 58% 20.3% 2.86* 42% 79.7% 0.53* Prof/Exec 46% 19.2% 2.40 54% 80.8% 0.67 $ > $75k 21% 9.5% 2.21 79% 90.5% 0.87 $ > $100k 7% 2.4% 2.92 93% 97.6% 0.95 * 58% / 20.3% * 42% / 79.7% 6 TIP When comparing customer profiles (Mike and Nancy), it is important to keep in mind the profile of the population as a whole. For this reason, using indexes (table #2) is often better than using raw values (table #1) Review Census Tract example on pages 94-95 7 Census Tract Example 8 Data Mining and Direct Marketing Campaigns Typical mailing of 100,000 pieces costs about $100,000 ($1/piece) Typical response rates < 10% Any list of prospects/customers that can be ranked by likelihood of response is good Campaign focused at top of list to increase response rate % 9 Consider the following 1,000,000 prospects Budget = $300,000 Mailing to 300,000 prospects Rank order list (model) vs no rank order: 0% 0% 100% 100% 30% 30% R E S P O N D E R S List Penetration 66% Benefit (66/30=2.17) No Model Model The ratio of concentration to penetration is the lift (2.17) (= model performance against no model). 10 Consider the following ROC chart / curve The false positive rate is plotted on the X- axis and one minus the false negative rate is plotted on the Y-axis. The area under the ROC curve is a measure of the models ability to differentiate between two outcomes. This measure is called discrimination. A perfect test has discrimination of 1 and a useless test for two outcomes has discrimination 0.5 since that is the area under the diagonal line that represents no model. 11 Consider the following Is the benefit worth the cost? Often, smaller, better-targeted campaign can be more profitable than a larger and more expensive one Be sure to consider real revenue (for example, 10 people buy = $100 revenue; 20 people buy = $200 revenue) Campaign profitability depends on many variables that can only be estimated, hence the need for an actual market test 12 Marketing Campaign Goal is to change behavior (to help drive revenue) How do we know if we did? Control Group randomly receives mailing Test Group model selected to get mailing Holdout Group model selected not get mailing Compare responses of the groups 13 Differential Response Analysis How do we know if the responders actually responded because of our campaign or would have anyway? Answer: Differential Response Analysis (DRA) reaching prospects who are more likely to make purchases because of having been contacted DRA starts with Control & Treated groups Control group = no mailing Treated group = receive mailing Compare resultssee if there is any uplift Control Group Treated Group Young Old Young Old Women 0.8% 0.4% 4.1% 4.6% Men 2.8% 3.3% 6.2% 5.2% 14 Differential Response Analysis 15 DM meets CRM* Matching campaigns to customers Segmenting the customer base Reducing exposure to credit risk Determining customer value Cross-selling and Up-selling Retention and Churn ([in]voluntary attrition) Different kinds of churn models predicting who will leave; predicting how long one will stay * Customer Relationship Management 16 RapidMiner Practice To see: Training Videos\01 - Ralf Klinkenberg RapidMinerResources\... 1 - Introduction -2- GUI Advanced.mp4 3 - Data Visualisation & Exploration -1- Introduction.mp4 3 - Data Visualisation & Exploration -4- Meta Data.mp4 To practice: Do the exercises presented in the movies using the files Iris.arff / Iris.ioo (RapidMiner data file) and Labor-Negociations.ioo.