Customer Relationship Management Through Data Mining

Customer Relationship Management Through Data Mining

BUDT 725: Models and Applications in Operations Research by Bruce L. Golden R.H. Smith School of Business Volume 1- Customer Relationship Management Through Data Mining 1 Customer Relationship Management Through Data Mining Introduction to Customer Relationship Management (CRM) Introduction to Data Mining Data Mining Software Churn Modeling Acquisition and Cross Sell Modeling 2 Relationship Marketing Relationship Marketing is a Process communicating with your customers listening to their responses Companies take actions marketing campaigns new products new channels new packaging 3 Relationship Marketing -- continued Customers and prospects respond most common response is no response

This results in a cycle data is generated opportunities to learn from the data and improve the process emerge 4 The Move Towards Relationship Management E-commerce companies want to customize the user experience Supermarkets want to be infomediaries Credit card companies want to recommend good restaurants and hotels in new cities Phone companies want to know your friends and family Bottom line: Companies want to be in the business of serving customers rather than merely selling products 5 CRM is Revolutionary Grocery stores have been in the business of stocking shelves Banks have been in the business of managing the spread between money borrowed and money lent Insurance companies have been in the business of managing loss ratios Telecoms have been in the business of completing telephone calls Key point: More companies are beginning to view customers as their primary asset 6

Why Now ? Representative Growth in a Maturing Market Number of Customers (1000s) 1200 total customers As growth flattens out, exploiting existing customers becomes more important 1000 800 600 In this region of rapid growth, building infrastructure is more important than CRM 400 new customers 200 churners 0 0 1

2 3 4 5 6 Year 7 8 9 10 11 7 The Electronic Trail A customer places a catalog order over the telephone At the local telephone company time of call, number dialed, long distance company used, At the long distance company (for the toll-free number) duration of call, route through switching system,

At the catalog items ordered, call center, promotion response, credit card used, inventory update, shipping method requested, 8 The Electronic Trail-- continued At the credit card clearing house transaction date, amount charged, approval code, vendor number, At the bank billing record, interest rate, available credit update, At the package carrier zip code, time stamp at truck, time stamp at sorting center, Bottom line: Companies do keep track of data 9 An Illustration A few years ago, UPS went on strike FedEx saw its volume increase After the strike, its volume fell FedEx identified those customers whose FedEx volumes had increased and then decreased These customers were using UPS again FedEx made special offers to these customers to get all of their business 10

The Corporate Memory Several years ago, Lands End could not recognize regular Christmas shoppers some people generally dont shop from catalogs but spend hundreds of dollars every Christmas if you only store 6 months of history, you will miss them Victorias Secret builds customer loyalty with a no-hassle returns policy some loyal customers return several expensive outfits each month they are really loyal renters 11 CRM Requires Learning and More Form a learning relationship with your customers Notice their needs On-line Transaction Processing Systems Remember their preferences Decision Support Data Warehouse Learn how to serve them better Data Mining Act to make customers more profitable 12 The Importance of Channels Channels are the way a company interfaces with

its customers Examples Direct mail Email Banner ads Telemarketing Billing inserts Customer service centers Messages on receipts Key data about customers come from channels 13 Channels -- continued Channels are the source of data Channels are the interface to customers Channels enable a company to get a particular message to a particular customer Channel management is a challenge in organizations CRM is about serving customers through all channels 14 Where Does Data Mining Fit In? Hindsight Analysis and Reporting (OLAP) Foresight Statistical Modeling

Insight Data Mining 15 Our Definition of Data Mining Exploration and analysis of large quantities of data By automatic or semi-automatic means To discover meaningful patterns and rules These patterns allow a company to better understand its customers improve its marketing, sales, and customer support operations Source: Berry and Linoff (1997) 16 Data Mining for Insight Classification Prediction Estimation Automatic Cluster Detection Affinity Grouping Description 17 Finding Prospects A cellular phone company wanted to introduce a new service They wanted to know which customers were the most likely prospects Data mining identified sphere of influence

as a key indicator of likely prospects Sphere of influence is the number of different telephone numbers that someone calls 18 Paying Claims A major manufacturer of diesel engines must also service engines under warranty Warranty claims come in from all around the world Data mining is used to determine rules for routing claims some are automatically approved others require further research Result: The manufacturer saves millions of dollars Data mining also enables insurance companies and the Fed. Government to save millions of dollars by not paying fraudulent medical insurance claims 19 Cross Selling Cross selling is another major application of data mining What is the best additional or best next offer (BNO) to make to each customer? E.g., a bank wants to be able to sell you automobile insurance when you get a car loan The bank may decide to acquire a full-service insurance agency 20 Holding on to Good Customers

Berry and Linoff used data mining to help a major cellular company figure out who is at risk for attrition And why are they at risk They built predictive models to generate call lists for telemarketing The result was a better focused, more effective retention campaign 21 Weeding out Bad Customers Default and personal bankruptcy cost lenders millions of dollars Figuring out who are your worst customers can be just as important as figuring out who are your best customers many businesses lose money on most of their customers 22 They Sometimes get Their Man The FBI handles numerous, complex cases such as the Unabomber case Leads come in from all over the country The FBI and other law enforcement agencies sift through thousands of reports from field agents looking for some connection Data mining plays a key role in FBI forensics 23 Anticipating Customer Needs Clustering is an undirected data mining technique

that finds groups of similar items Based on previous purchase patterns, customers are placed into groups Customers in each group are assumed to have an affinity for the same types of products New product recommendations can be generated automatically based on new purchases made by the group This is sometimes called collaborative filtering 24 CRM Focuses on the Customer The enterprise has a unified view of each customer across all business units and across all channels This is a major systems integration task The customer has a unified view of the enterprise for all products and regardless of channel This requires harmonizing all the channels 25 A Continuum of Customer Relationships Large accounts have sales managers and account teams E.g., Coca-Cola, Disney, and McDonalds CRM tends to focus on the smaller customer -the consumer But, small businesses are also good candidates for CRM 26

What is a Customer A transaction? An account? An individual? A household? The customer as a transaction purchases made with cash are anonymous most Web surfing is anonymous we, therefore, know little about the consumer 27 A Customer is an Account More often, a customer is an account Retail banking checking account, mortgage, auto loan, Telecommunications long distance, local, ISP, mobile, Insurance auto policy, homeowners, life insurance, Utilities The account-level view of a customer also misses the boat since each customer can have multiple accounts 28 Customers Play Different Roles Parents buy back-to-school clothes for teenage

children children decide what to purchase parents pay for the clothes parents own the transaction Parents give college-age children cellular phones or credit cards parents may make the purchase decision children use the product It is not always easy to identify the customer 29 The Customers Lifecycle Childhood birth, school, graduation, Young Adulthood choose career, move away from parents, Family Life marriage, buy house, children, divorce, Retirement sell home, travel, hobbies, Much marketing effort is directed at each stage of life 30 The Customers Lifecycle is Unpredictable It is difficult to identify the appropriate events graduation, retirement may be easy

marriage, parenthood are not so easy many events are one-time Companies miss or lose track of valuable information a man moves a woman gets married, changes her last name, and merges her accounts with spouse It is hard to track your customers so closely, but, to the extent that you can, many marketing opportunities arise 31 Customers Evolve Over Time Customers begin as prospects Prospects indicate interest fill out credit card applications apply for insurance visit your website They become new customers After repeated purchases or usage, they become established customers Eventually, they become former customers either voluntarily or involuntarily 32 Business Processes Organize Around the Customer Lifecycle Acquisition Activation

Relationship Management Winback Former Customer High Value Prospect New Customer Established Customer High Potential Low Value Voluntary Churn Forced Churn 33 Different Events Occur Throughout the Lifecycle Prospects receive marketing messages When they respond, they become new customers

They make initial purchases They become established customers and are targeted by cross-sell and up-sell campaigns Some customers are forced to leave (cancel) Some leave (cancel) voluntarily Others simply stop using the product (e.g., credit card) 34 Winback/collection campaigns Different Data is Available Throughout the Lifecycle The purpose of data warehousing is to keep this data around for decision-support purposes Charles Schwab wants to handle all of their customers investment dollars Schwab observed that customers started with small investments 35 Different Data is Available Throughout the Lifecycle -- continued By reviewing the history of many customers, Schwab discovered that customers who transferred large amounts into their Schwab accounts did so soon after joining After a few months, the marketing cost could not be justified Schwabs marketing strategy changed as a result 36 Different Models are Appropriate

at Different Stages Prospect acquisition Prospect product propensity Best next offer Forced churn Voluntary churn Bottom line: We use data mining to predict certain events during the customer lifecycle 37 Different Approaches to Data Mining Outsourcing let an outside expert do the work have him/her report the results Off-the-shelf, turn-key software solutions packages have generic churn models & response models they work pretty well Master Data Mining develop expertise in-house use sophisticated software such as Clementine or Enterprise Miner 38 Privacy is a Serious Matter Data mining and CRM raise some privacy concerns These concerns relate to the collection of data, more than the analysis of data

The next few slides illustrate marketing mistakes that can result from the abundance and availability of data 39 Using Data Mining to Help Diabetics Early detection of diabetes can save money by preventing more serious complications Early detection of complications can prevent worsening retinal eye exams every 6 or 12 months can prevent blindness these eye exams are relatively inexpensive So one HMO took action they decided to encourage their members, who had diabetes to get eye exams the IT group was asked for a list of members with diabetes 40 One Womans Response Letters were sent out to HMO members Three types of diabetes congenital, adult-onset, gestational One woman contacted had gestational diabetes several years earlier She was traumatized by the letter, thinking the diabetes had recurred She threatened to sue the HMO Mistake: Disconnect between the domain expertise and data expertise 41

Gays in the Military The dont ask; dont tell policy allows discrimination against openly gay men and lesbians in the military Identification as gay or lesbian is sufficient grounds for discharge This policy is enforced Approximately 1000 involuntary discharges each year 42 The Story of Former Senior Chief Petty Officer Timothy McVeigh Several years ago, McVeigh used an AOL account, with an anonymous alias Under marital status, he listed gay A colleague discovered the account and called AOL to verify that the owner was McVeigh AOL gave out the information over the phone McVeigh was discharged (three years short of his pension) The story doesnt end here 43 Two Serious Privacy Violations AOL breached its own policy by giving out confidential user information AOL paid an undisclosed sum to settle with McVeigh and suffered bad press as well The law requires that government agents identify themselves to get online subscription information

This was not done McVeigh received an honorable discharge with full retirement pension 44 Friends, Family, and Others In the 1990s, MCI promoted the Friends and Family program They asked existing customers for names of people they talked with often If these friends and family signed up with MCI, then calls to them would be discounted Did MCI have to ask customers about who they call regularly? Early in 1999, BT (formerly British Telecom) took the idea one step beyond BT invented a new marketing program discounts to the most frequently called numbers 45 BT Marketing Program BT notified prospective customers of this program by sending them their most frequently called numbers One woman received the letter uncovered her husbands cheating threw him out of the house sued for divorce The husband threatened to sue BT for violating his privacy

BT suffered negative publicity 46 No Substitute for Human Intelligence Data mining is a tool to achieve goals The goal is better service to customers Only people know what to predict Only people can make sense of rules Only people can make sense of visualizations Only people know what is reasonable, legal, tasteful Human decision makers are critical to the data mining process 47 A Long, Long Time Ago There was no marketing There were few manufactured goods Distribution systems were slow and uncertain There was no credit Most people made what they needed at home There were no cell phones There was no data mining It was sufficient to build a quality product and get it to market 48 Then and Now Before supermarkets, a typical grocery store carried 800 different items A typical grocery store today carries tens of thousands of different items There is intense competition for shelf space and

premium shelf space In general, there has been an explosion in the number of products in the last 50 years Now, we need to anticipate and create demand (e.g., e-commerce) This is what marketing is all about 49 Effective Marketing Presupposes High quality goods and services Effective distribution of goods and services Adequate customer service Marketing promises are kept Competition direct (same product) wallet-share Ability to interact directly with customers 50 The ACME Corporation Imagine a fictitious corporation that builds widgets It can sell directly to customers via a catalog or the Web maintain control over brand and image It can sell directly through retail channels get help with marketing and advertising It can sell through resellers outsource marketing and advertising entirely Lets assume ACME takes the direct marketing

approach 51 Before Focusing on One-to-One Marketing Branding is very important provides a mark of quality to consumers old concept Bordeaux wines, Chinese porcelain, Bruges cloth really took off in the 20th Century Advertising is hard media mix problem print, radio, TV, billboard, Web difficult to measure effectiveness Half of my advertising budget is wasted; I just dont know which half. 52 Different Approaches to Direct Marketing Nave Approach get a list of potential customers send out a large number of messages and repeat Wave Approach send out a large number of messages and test Staged Approach send a series of messages over time Controlled Approach send out messages over time to control response (e.g., get 10,000 responses/week) 53

The World is Speeding Up Advertising campaigns take months market research design and print material Catalogs are planned seasons in advance Direct mail campaigns also take months Telemarketing campaigns take weeks Web campaigns take days modification/refocusing is easy 54 How Data Mining Helps in Marketing Campaigns Improves profit by limiting campaign to most likely responders Reduces costs by excluding individuals least likely to respond AARP mails an invitation to those who turn 50 they excluded the bottom 10% of their list response rate did not suffer 55 How Data Mining Helps in Marketing Campaigns--continued Predicts response rates to help staff call centers, with inventory control, etc. Identifies most important channel for each customer Discovers patterns in customer data

56 Some Background on ACME They are going to pursue a direct marketing approach Direct mail marketing budget is $300,000 Best estimates indicate between 1 and 10 million customers ACME wants to target the customer base costeffectively ACME seeks to assign a score to each customer which reflects the relative likelihood of that customer purchasing the product 57 How Do You Assign Scores Randomly everyone gets the same score Assign relative scores based on ad-hoc business knowledge Assign a score to each cell in an RFM (recency, frequency, monetary) analysis Profile existing customers and use these profiles to assign scores to similar, potential customers Build a predictive model based on similar product sales in the past 58 Data Mining Models Assign a Score to Each Customer ID 0102 0104 0105

0110 0111 0112 0116 0117 0118 Name Will Sue John Lori Beth Pat David Frank Ethel State MA NY AZ AZ NM WY ID MS NE Score Rank 0.314 7 0.159

9 0.265 8 0.358 5 0.979 1 0.328 6 0.446 4 0.897 2 0.446 4 Comments 1. Think of score as likelihood of responding 2. Some scores may be the same 59 Approach 1: Budget Optimization ACME has a budget of $300,000 for a direct mail campaign Assumptions each item being mailed costs $1 this cost assumes a minimum order of 20,000 ACME can afford to contact 300,000 customers

ACME contacts the highest scoring 300,000 customers Lets assume ACME is selecting from the top three deciles 60 The Concept of Lift If we look at a random 10% of the potential customers, we expect to get 10% of likely responders Can we select 10% of the potential customers and get more than 10% of likely responders? If so, we realize lift This is a key goal in data mining 61 The Cumulative Gains Chart Notes 100% 90% 80% 1. x-axis gives population percentile 2. y-axis gives %

captured responses 3. customers with top 10% of the scores account for 30% of responders 70% 60% 50% 40% 30% 20% Response M odel 10% No M odel 0% 0% 10% 20% 30%

40% 50% 60% 70% 80% 90% 100% 62 The Actual Lift Chart 3.5 Notes 1. x-axis gives population percentile 2 2. y-axis gives the lift 1.5

3. the top 10% of the scorers are 3 times more likely to respond than a random 10% would be 3 2.5 1 Response M odel 0.5 No M odel 0 10% 20% 30% 40% 50% 60% 70%

80% 90% 100% 63 How Well Does ACME Do? ACME selects customers from the top three deciles From cumulative gains chart, a response rate of 65% (vs. 30%) results From lift chart, we see a lift of 65/30 = 2.17 The two charts convey the same information, but in different ways 64 Can ACME Do Better? Test marketing campaign send a mailing to a subset of the customers, say 30,000 take note of the 1 to 2% of those who respond build predictive models to predict response use the results from these models The key is to learn from the test marketing campaign 65 Optimizing the Budget Decide on the budget

Based on cost figures, determine the size of the mailing Develop a model to score all customers with respect to their relative likelihood to respond to the offer Choose the appropriate number of top scoring customers 66 Approach 2: Optimizing the Campaign Lift allows us to contact more of the potential responders It is a very useful measure But, how much better off are we financially? We seek a profit-and-loss statement for the campaign To do this, we need more information than before 67 Is the Campaign Profitable? Suppose the following the typical customer will purchase about $100 worth of merchandise from the next catalog of the $100, $55 covers the cost of inventory, warehousing, shipping, and so on the cost of sending mail to each customer is $1 Then, the net revenue per customer in the campaign is $100 - $55 - $1 = $44 68 The Profit/Loss Matrix

Those predicted to respond cost $1 those who actually respond yield a gain of $45 those who dont respond yield no gain Predicted Someone who scores in the top 30%, is predicted to respond ACTUAL YES NO YES $44 -$1 NO $0 $0 Those not predicted to respond cost $0 and yield no gain 69

The Profit/Loss Matrix--continued The profit/loss matrix is a powerful concept But, it has its limitations people who dont respond become more aware of the brand/product due to the marketing campaign they may respond next time people not contacted might have responded had they been invited For now, lets focus on the profit/loss matrix 70 How Do We Get the P/L Numbers? Cost numbers are relatively easy mailing and printing costs can be handled by accounts payable call center costs, for incoming orders, are usually fixed Revenue numbers are rough estimates based on previous experience, back-of-envelope calculations, guesswork based on models of customer buying behavior 71 Is the Campaign Profitable? Assumptions made so far $44 net revenue per responder ($1) net revenue per non-responder 300,000 in target group new assumption: overhead charge of $20,000

Resulting lift is 2.17 We can now estimate profit for different response rates 72 Net Revenue for the Campaign Resp Rate Overall Resp Rate Size (YES) Size (NO) Total Net Revenue 1% 0.46% 3,000 297,000 300,000 $ (185,000) 2%

0.92% 6,000 294,000 300,000 $ (50,000) 3% 1.38% 9,000 291,000 300,000 $ 85,000 4% 1.85% 12,000 288,000 300,000 $ 220,000

5% 2.31% 15,000 285,000 300,000 $ 355,000 6% 2.77% 18,000 282,000 300,000 $ 490,000 7% 3.23% 21,000 279,000

300,000 $ 625,000 8% 3.69% 24,000 276,000 300,000 $ 760,000 9% 4.15% 27,000 273,000 300,000 $ 895,000 10% 4.62% 30,000

270,000 300,000 $ 1,030,000 The campaign makes money if it achieves a response rate of at least 3% 73 Net Revenue Table Explained Suppose response rate of 3% Net revenue = 9000 44 + 291,000 (-1) - 20,000 = $85,000 Lift = response rate for campaign overall response rate overall response rate = response rate for campaign lift 3% = 2.17 = 1.38% Suppose response rate of 6% 74 Two Ways to Estimate Response Rates Use a randomly selected hold-out set (the test set) this data is not used to build the model the models performance on this set estimates the performance on unseen data Use a hold-out set on oversampled data most data mining involves binary outcomes

often, we try to predict a rare event (e.g., fraud) with oversampling, we overrepresent the rare outcomes and underrepresent the common outcomes 75 Oversampling Builds Better Models for Rare Events Suppose 99% of records involve no fraud A model that always predicts no fraud will be hard to beat But, such a model is not useful Stratified sampling with two outcomes is called oversampling 76 Return to Earlier Model 100% DECILE GAINS 0% 0% 0% 0.000 80%

10% 30% 30% 3.000 70% 20% 20% 50% 2.500 60% 30% 15% 65% 2.167 40% 13%

78% 1.950 50% 7% 85% 1.700 60% 5% 90% 1.500 70% 4% 94% 1.343 80% 4% 98%

1.225 90% 2% 100% 1.111 100% 0% 100% 1.000 90% 50% 40% 30% 20% Response Model 10% No Model 0%

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% CUM LIFT 100% 77 Assume an Overall Response Rate of 1% and Calculate the Profit for Each Decile DECILE GAINS

CUM LIFT SIZE SIZE(YES) SIZE(NO) PROFIT - - ($20,000) 0% 0.00% 0% 0.000 0 10% 30.00%

30% 3.000 100,000 3,000 97,000 $15,000 20% 20.00% 50% 2.500 200,000 5,000 195,000 $5,000 30% 15.00% 65%

2.167 300,000 6,500 293,500 ($27,500) 40% 13.00% 78% 1.950 400,000 7,800 392,200 ($69,000) 50% 7.00% 85%

1.700 500,000 8,500 491,500 ($137,500) 60% 5.00% 90% 1.500 600,000 9,000 591,000 ($215,000) 70% 4.00% 94% 1.343

700,000 9,400 690,600 ($297,000) 80% 4.00% 98% 1.225 800,000 9,800 790,200 ($397,000) 90% 2.00% 100% 1.111

900,000 10,000 890,000 ($470,000) 100% 0.00% 100% 1.000 1,000,000 10,000 990,000 ($570,000) Remember: $44 net revenue/ $1 cost per item mailed/ $20,000 overhead 78 Review of Profit Calculation Key equations sizelift size (yes) =

profit = 44 100 size (yes) - size (no) - 20,000 Example: top three deciles (30% row) size (yes) = 2.167 = 6500 300,000 profit = 286,000 100 - 293,500 - 20,000 = -27,500 Notice that top 10% yields the maximum profit Mailing to the top three deciles would cost us money 79 Typical Shape for a Profit Curve ($44, $1, $20,000) PROFIT by Decile 100,000 0 -100,000 -200,000 -300,000 -400,000 -500,000 -600,000

-700,000 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 80 Approach 2 Summary Estimate cost per contact, overhead, and estimated revenue per responder Build a model and estimate response probabilities for each customer Order the customers by their response scores For each decile, calculate the cumulative number of responders and non-responders

Using the estimates, determine the cumulative profit for each decile Choose all the deciles up to the one with the highest cumulative profit 81 The Problem with Campaign Optimization Campaign optimization is very sensitive to the underlying assumptions Suppose the response rate is 2% rather than 1%? Suppose the cost of contacting a customer is $1.20 rather than $1? Sensitivity is a serious problem 82 Assume an Overall Response Rate of 1.2% and Calculate the Profit for Each Decile DECILE GAINS CUM LIFT SIZE SIZE(YES) SIZE(NO) PROFIT

- - ($20,000) 0% 0.00% 0% 0.000 0 10% 30.00% 30% 3.000 100,000 3,600 96,400 $42,000

20% 20.00% 50% 2.500 200,000 6,000 194,000 $50,000 30% 15.00% 65% 2.167 300,000 7,801 292,199 $31,054 40%

13.00% 78% 1.950 400,000 9,360 390,640 $1,200 50% 7.00% 85% 1.700 500,000 10,200 489,800 ($61,000) 60%

5.00% 90% 1.500 600,000 10,800 589,200 ($134,000) 70% 4.00% 94% 1.343 700,000 11,281 688,719 ($212,346) 80% 4.00%

98% 1.225 800,000 11,760 788,240 ($290,800) 90% 2.00% 100% 1.111 900,000 11,999 888,001 ($380,054) 100% 0.00%

100% 1.000 1,000,000 12,000 988,000 ($480,000) Remember: $44 net revenue/ $1 cost per item mailed/ $20,000 overhead 83 Assume an Overall Response Rate of 0.8% and Calculate the Profit for Each Decile DECILE GAINS CUM LIFT SIZE SIZE(YES) SIZE(NO) PROFIT

- - ($20,000) 0% 0.00% 0% 0.000 0 10% 30.00% 30% 3.000 100,000 2,400 97,600 ($12,000) 20%

20.00% 50% 2.500 200,000 4,000 196,000 ($40,000) 30% 15.00% 65% 2.167 300,000 5,201 294,799 ($85,964) 40%

13.00% 78% 1.950 400,000 6,240 393,760 ($139,200) 50% 7.00% 85% 1.700 500,000 6,800 493,200 ($214,000) 60% 5.00%

90% 1.500 600,000 7,200 592,800 ($296,000) 70% 4.00% 94% 1.343 700,000 7,521 692,479 ($381,564) 80% 4.00%

98% 1.225 800,000 7,840 792,160 ($467,200) 90% 2.00% 100% 1.111 900,000 7,999 892,001 ($560,036) 100% 0.00% 100%

1.000 1,000,000 8,000 992,000 ($660,000) Remember: $44 net revenue/ $1 cost per item mailed/ $20,000 overhead 84 Assume an Overall Response Rate of 2% and Calculate the Profit for Each Decile DECILE GAINS CUM LIFT SIZE SIZE(YES) SIZE(NO) PROFIT -

- ($20,000) 0% 0.00% 0% 0.000 0 10% 30.00% 30% 3.000 100,000 6,000 94,000 $150,000 20%

20.00% 50% 2.500 200,000 10,000 190,000 $230,000 30% 15.00% 65% 2.167 300,000 13,002 286,998 $265,090 40% 13.00%

78% 1.950 400,000 15,600 384,400 $282,000 50% 7.00% 85% 1.700 500,000 17,000 483,000 $245,000 60% 5.00%

90% 1.500 600,000 18,000 582,000 $190,000 70% 4.00% 94% 1.343 700,000 18,802 681,198 $126,090 80% 4.00% 98%

1.225 800,000 19,600 780,400 $62,000 90% 2.00% 100% 1.111 900,000 19,998 880,002 ($20,090) 100% 0.00% 100%

1.000 1,000,000 20,000 980,000 ($120,000) Remember: $44 net revenue/ $1 cost per item mailed/ $20,000 overhead 85 Dependence on Response Rate ($44, $1, $20,000) $400,000 $200,000 $0 ($200,000) ($400,000) PROF (1.00%) PROF (1.20%) ($600,000) PROF (0.80%) PROF (2.00%) ($800,000) 0% 10%

20% 30% 40% 50% 60% 70% 80% 90% 100% 86 Assume an Overall Response Rate of 1% and Calculate the Profit for Each Decile DECILE GAINS CUM LIFT SIZE SIZE(YES)

SIZE(NO) PROFIT - - ($20,000) 0% 0.00% 0% 0.000 0 10% 30.00% 30% 3.000 100,000 3,000

97,000 ($4,400) 20% 20.00% 50% 2.500 200,000 5,000 195,000 ($34,000) 30% 15.00% 65% 2.167 300,000 6,500 293,500

($86,200) 40% 13.00% 78% 1.950 400,000 7,800 392,200 ($147,440) 50% 7.00% 85% 1.700 500,000 8,500 491,500

($235,800) 60% 5.00% 90% 1.500 600,000 9,000 591,000 ($333,200) 70% 4.00% 94% 1.343 700,000 9,400 690,600 ($435,120)

80% 4.00% 98% 1.225 800,000 9,800 790,200 ($537,040) 90% 2.00% 100% 1.111 900,000 10,000 890,000 ($648,000)

100% 0.00% 100% 1.000 1,000,000 10,000 990,000 ($768,000) Remember: $44 net revenue/ $1.2 cost per item mailed/ $20,000 overhead 87 Assume an Overall Response Rate of 1% and Calculate the Profit for Each Decile DECILE GAINS CUM LIFT SIZE SIZE(YES)

SIZE(NO) PROFIT - - ($20,000) 0% 0.00% 0% 0.000 0 10% 30.00% 30% 3.000 100,000 3,000

97,000 $34,400 20% 20.00% 50% 2.500 200,000 5,000 195,000 $44,000 30% 15.00% 65% 2.167 300,000 6,500 293,500

$31,200 40% 13.00% 78% 1.950 400,000 7,800 392,200 $9,440 50% 7.00% 85% 1.700 500,000 8,500 491,500

($39,200) 60% 5.00% 90% 1.500 600,000 9,000 591,000 ($96,800) 70% 4.00% 94% 1.343 700,000 9,400 690,600 ($158,880)

80% 4.00% 98% 1.225 800,000 9,800 790,200 ($220,000) 90% 2.00% 100% 1.111 900,000 10,000 890,000 ($292,000)

100% 0.00% 100% 1.000 1,000,000 10,000 990,000 ($372,000) Remember: $44 net revenue/ $0.8 cost per item mailed/ $20,000 overhead 88 Assume an Overall Response Rate of 1% and Calculate the Profit for Each Decile DECIL E GAINS CUM LIFT SIZE SIZE(YES)

SIZE(NO) PROFIT - - ($20,000) 0% 0.00% 0% 0.000 0 10% 30.00% 30% 3.000 100,000 3,000

97,000 ($82,000) 20% 20.00% 50% 2.500 200,000 5,000 195,000 ($190,000) 30% 15.00% 65% 2.167 300,000 6,500 293,500

($321,000) 40% 13.00% 78% 1.950 400,000 7,800 392,200 ($461,200) 50% 7.00% 85% 1.700 500,000 8,500 491,500

($629,000) 60% 5.00% 90% 1.500 600,000 9,000 591,000 ($806,000) 70% 4.00% 94% 1.343 700,000 9,400 690,600 ($987,600)

80% 4.00% 98% 1.225 800,000 9,800 790,200 ($1,169,200) 90% 2.00% 100% 1.111 900,000 10,000 890,000 ($1,360,000)

100% 0.00% 100% 1.000 1,000,000 10,000 990,000 ($1,560,000) Remember: $44 net revenue/ $2 cost per item mailed/ $20,000 overhead 89 Dependence on Costs $200,000 $0 ($200,000) ($400,000) ($600,000) ($800,000) ($1,000,000) PROF ($1) ($1,200,000) PROF ($1.20)

($1,400,000) PROF ($0.80) ($1,600,000) PROF ($2.00) ($1,800,000) 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 90

Assume an Overall Response Rate of 1% and Calculate the Profit for Each Decile DECILE GAINS CUM LIFT SIZE SIZE(YES) SIZE(NO) PROFIT - - ($20,000) 0% 0.00% 0% 0.000 0

10% 30.00% 30% 3.000 100,000 3,000 97,000 ($11,400) 20% 20.00% 50% 2.500 200,000 5,000 195,000 ($39,000)

30% 15.00% 65% 2.167 300,000 6,500 293,500 ($84,700) 40% 13.00% 78% 1.950 400,000 7,800 392,200 ($137,640) 50%

7.00% 85% 1.700 500,000 8,500 491,500 ($212,300) 60% 5.00% 90% 1.500 600,000 9,000 591,000 ($294,200) 70%

4.00% 94% 1.343 700,000 9,400 690,600 ($379,720) 80% 4.00% 98% 1.225 800,000 9,800 790,200 ($465,240) 90% 2.00%

100% 1.111 900,000 10,000 890,000 ($558,000) 100% 0.00% 100% 1.000 1,000,000 10,000 990,000 ($658,000) Remember: $35.2 net revenue/ $1 cost per item mailed/ $20,000 overhead 91 Assume an Overall Response Rate of 1%

and Calculate the Profit for Each Decile DECILE GAINS CUM LIFT SIZE SIZE(YES) SIZE(NO) PROFIT - - ($20,000) 0% 0.00% 0% 0.000 0

10% 30.00% 30% 3.000 100,000 3,000 97,000 $41,400 20% 20.00% 50% 2.500 200,000 5,000 195,000 $49,000 30%

15.00% 65% 2.167 300,000 6,500 293,500 $29,700 40% 13.00% 78% 1.950 400,000 7,800 392,200 ($360) 50%

7.00% 85% 1.700 500,000 8,500 491,500 ($62,700) 60% 5.00% 90% 1.500 600,000 9,000 591,000 ($135,800) 70% 4.00%

94% 1.343 700,000 9,400 690,600 ($214,280) 80% 4.00% 98% 1.225 800,000 9,800 790,200 ($292,760) 90% 2.00%

100% 1.111 900,000 10,000 890,000 ($382,000) 100% 0.00% 100% 1.000 1,000,000 10,000 990,000 ($482,000) Remember: $52.8 net revenue/ $1 cost per item mailed/ $20,000 overhead 92 Assume an Overall Response Rate of 1% and Calculate the Profit for Each Decile

DECILE GAINS CUM LIFT SIZE SIZE(YES) SIZE(NO) PROFIT - - ($20,000) 0% 0.00% 0% 0.000 0 10%

30.00% 30% 3.000 100,000 3,000 97,000 $147,000 20% 20.00% 50% 2.500 200,000 5,000 195,000 $225,000 30%

15.00% 65% 2.167 300,000 6,500 293,500 $258,500 40% 13.00% 78% 1.950 400,000 7,800 392,200 $274,200 50% 7.00%

85% 1.700 500,000 8,500 491,500 $236,500 60% 5.00% 90% 1.500 600,000 9,000 591,000 $181,000 70% 4.00%

94% 1.343 700,000 9,400 690,600 $116,600 80% 4.00% 98% 1.225 800,000 9,800 790,200 $52,200 90% 2.00% 100%

1.111 900,000 10,000 890,000 ($30,000) 100% 0.00% 100% 1.000 1,000,000 10,000 990,000 ($130,000) Remember: $88 net revenue/ $1 cost per item mailed/ $20,000 overhead 93 Dependence on Revenue $400,000 $200,000

$0 ($200,000) ($400,000) PROF ($44) PROF ($35.20) ($600,000) PROF ($52.80) PROF ($88) ($800,000) 0% 10% 20% 30% 40% 50% 60% 70% 80% 90%

100% 94 Campaign Optimization Drawbacks Profitability depends on response rates, cost estimates, and revenue potential Each one impacts profitability The numbers we use are just estimates If we are off by a little here and a little there, our profit estimates could be off by a lot In addition, the same group of customers is chosen for multiple campaigns 95 Approach 3: Customer Optimization Campaign optimization makes a lot of sense But, campaign profitability is difficult to estimate Is there a better way? Do what is best for each customer Focus on customers, rather than campaigns 96 Real-World Campaigns Companies usually have several products that they want to sell telecom: local, long distance, mobile, ISP, etc. banking: CDs, mortgages, credit cards, etc. insurance: home, car, personal liability, etc. retail: different product lines There are also upsell and customer retention programs

These campaigns compete for customers 97 Each Campaign May Have a Separate Model These models produce scores The score tells us how likely a given customer is to respond to that specific campaign 0, if the customer already has the product 0, if the product and customer are incompatible 1, if the customer has asked about the product Each campaign is relevant for a subset of all the customers Imagine three marketing campaigns, each with a separate data mining model 98 Sample Scores (as Rankings for Three Different Campaigns) ID Name State Mod A 0102 0104 0105 0110 0111 0112 0116 0117 0118

Will Sue John Lori Beth Pat David Frank Ethel MA NY AZ AZ NM WY ID MS NE 3 1 2 5 9 4 6 8 6 Mod B

Mod C 4 2 1 7 3 5 5 9 8 2 4 1 6 8 2 7 8 5 99 Choose the Best Customers for Each Campaign ID 0102 0104 0105 0110 0111 0112 0116 0117

0118 Name Will Sue John Lori Beth Pat David Frank Ethel State MA NY AZ AZ NM WY ID MS NE Mod A Mod B Mod C 3 4 2 1 2 4 2 1

1 5 7 6 9 3 8 4 5 2 6 5 7 8 9 8 6 8 5 100 A Common Situation Good customers are typically targeted by many campaigns Many other customers are not chosen for any campaigns Good customers who become inundated with contacts become less likely to respond at all Let the campaigns compete for customers 101 Choose the Best Campaign for Each Customer

ID 0102 0104 0105 0110 0111 0112 0116 0117 0118 Name Will Sue John Lori Beth Pat David Frank Ethel State MA NY AZ AZ NM WY ID MS NE

Mod A Mod B Mod C 3 4 2 1 2 4 2 1 1 5 7 6 9 3 8 4 5 2 6 5 7 8 9 8 6 8 5 102 Focus on the Customer Determine the propensity of each customer to respond to each campaign

Estimate the net revenue for each customer from each campaign Incorporate profitability into the customeroptimization strategy Not all campaigns will apply to all customers 103 First, Determine Response Rate for Each Campaign ID Name State Mod A Mod B Mod C 102 Will MA 2.0% 0.9% 0.0% 104

Sue NY 8.0% 1.4% 3.7% 105 John AZ 3.8% 2.3% 11.0% 110 Lori AZ 0.9% 7.0%

1.3% 111 Beth NM 0.1% 1.2% 0.8% 112 Pat WY 2.0% 0.8% 4.6% 116 David ID 0.8%

0.8% 1.1% 117 Frank MS 0.2% 0.2% 0.8% 118 Ethel NE 0.8% 0.2% 0.0% Customers who are not candidates are given a rate of zero104 Second, Add in Product Profitability ID

Name State Mod A Mod B 102 Will 104 Mod C Prof A Prof B Prof C MA 2.0% 0.9% 0.0% $56 $72 $20 Sue NY

8.0% 1.4% 3.7% $56 $72 $20 105 John AZ 3.8% 2.3% 11.0% $56 $72 $20 110 Lori

AZ 0.9% 7.0% 1.3% $56 $72 $20 111 Beth NM 0.1% 1.2% 0.8% $56 $72 $20

112 Pat WY 2.0% 0.8% 4.6% $56 $72 $20 116 David ID 0.8% 0.8% 1.1% $56 $72

$20 117 Frank MS 0.2% 0.2% 0.8% $56 $72 $20 118 Ethel NE 0.8% 0.2% 0.0%

$56 $72 $20 As a more sophisticated alternative, profit could be estimated for each customer/product combination 105 Finally, Determine the Campaign with the Highest Value ID Name State EP (A) EP (B) EP (C) Campaign 102 Will MA $1.12

$0.65 $0.00 A 104 Sue NY $4.48 $1.01 $0.74 A 105 John AZ $2.13 $1.66 $2.20

C 110 Lori AZ $0.50 $5.04 $0.26 B 111 Beth NM $0.06 $0.86 $0.16 B 112 Pat

WY $1.12 $0.58 $0.92 A 116 David ID $0.45 $ 0.58 $0.22 B 117 Frank MS $0.11

$0.14 $0.16 C 118 Ethel NE $0.45 $0.14 $0.00 A EP (k) = the expected profit of product k For each customer, choose the highest expected profit campaign 106 Conflict Resolution with Multiple Campaigns Managing many campaigns at the same time is complex for technical and political reasons Who owns the customer? Handling constraints

each campaign is appropriate for a subset of customers each campaign has a minimum and maximum number of contacts each campaign seeks a target response rate new campaigns emerge over time 107 Marketing Campaigns and CRM The simplest approach is to optimize the budget using the rankings that models produce Campaign optimization determines the most profitable subset of customers for a given campaign, but it is sensitive to assumptions Customer optimization is more sophisticated It chooses the most profitable campaign for each customer 108 The Data Mining Process What role does data mining play within an organization? How does one do data mining correctly? The SEMMA Process select and sample explore modify model assess 109 Identify the Right Business Problem Involve the business users Have them provide business expertise, not

technical expertise Define the problem clearly predict the likelihood of churn in the next month for our 10% most valuable customers Define the solution clearly is this a one-time job, an on-going monthly batch job, or a real-time response (call centers and web)? What would the ideal result look like? how would it be used? 110 Transforming the Data into Actionable Information Select and sample by extracting a portion of a large data set-- big enough to contain significant information, but small enough to manipulate quickly Explore by searching for unanticipated trends and anomalies in order to gain understanding 111 Transforming the Data into Actionable Information-- continued Modify by creating, selecting, and transforming the variables to focus the model selection process Model by allowing the software to search automatically for a combination of variables that reliably predicts a desired outcome Assess by evaluating the usefulness and reliability of the findings from the data mining

process 112 Act on Results Marketing/retention campaign lists or scores Personalized messages Customized user experience Customer prioritization Increased understanding of customers, products, messages 113 Measure the Results Confusion matrix Cumulative gains chart Lift chart Estimated profit 114 Data Mining Uses Data from the Past to Effect Future Action Those who do not remember the past are condemned to repeat it. George Santayana Analyze available data (from the past) Discover patterns, facts, and associations Apply this knowledge to future actions

115 Examples Prediction uses data from the past to make predictions about future events (likelihoods and probabilities) Profiling characterizes past events and assumes that the future is similar to the past (similarities) Description and visualization find patterns in past data and assume that the future is similar to the past 116 We Want a Stable Model A stable model works (nearly) as well on unseen data as on the data used to build it Stability is more important than raw performance for most applications we want a car that performs well on real roads, not just on test tracks Stability is a constant challenge 117 Is the Past Relevant? Does past data contain the important business drivers? e.g., demographic data Is the business environment from the past relevant to the future? in the ecommerce era, what we know about the past

may not be relevant to tomorrow users of the web have changed since late 1990s Are the data mining models created from past data relevant to the future? have critical assumptions changed? 118 Data Mining is about Creating Models A model takes a number of inputs, which often come from databases, and it produces one or more outputs Sometimes, the purpose is to build the best model The best model yields the most accurate output Such a model may be viewed as a black box Sometimes, the purpose is to better understand what is happening This model is more like a gray box 119 Models Past Present Data ends here Future Actions take place here

Building models takes place in the present using data from the past outcomes are already known Applying (or scoring) models takes place in the present Acting on the results takes place in the future outcomes are not known 120 Often, the Purpose is to Assign a Score to Each Customer ID Name State Score 1 0102 Will MA 0.314 2 0104 Sue NY

0.159 3 0105 John AZ 0.265 4 0110 Lori AZ 0.358 5 0111 Beth NM 0.979

6 0112 Pat WY 0.328 7 0116 David ID 0.446 8 0117 Frank MS 0.897 9 0118

Ethel NE 0.446 Comments 1. Scores are assigned to rows using models 2. Some scores may be the same 3. The scores may represent the probability of some outcome 121 Common Examples of What a Score Could Mean Likelihood to respond to an offer Which product to offer next Estimate of customer lifetime Likelihood of voluntary churn Likelihood of forced churn Which segment a customer belongs to Similarity to some customer profile Which channel is the best way to reach the customer 122 The Scores Provide a Ranking of the Customers ID

Name State Score ID 1 0102 Will MA 0.314 5 0111 2 0104 Sue NY 0.159 8

0117 3 0105 John AZ 0.265 4 0110 Lori AZ 0.358 5 0111 Beth NM 6 0112

Pat 7 Name State Score Beth NM 0.979 MS 0.897 Frank SORT 7 0116 David ID 0.446 0.979 9

0118 Ethel NE 0.446 WY 0.328 4 0110 Lori AZ 0.358 0116 David ID 0.446 6 0112

Pat WY 0.328 8 0117 Frank MS 0.897 1 0102 Will MA 0.314 9 0118 Ethel NE

0.446 3 0105 John AZ 0.265 2 0104 Sue NY 0.159 123 This Ranking give Rise to Quantiles (terciles, quintiles, deciles, etc.) ID Name State Score

5 0111 Beth NM 0.979 8 0117 Frank MS 0.897 7 0116 David ID 0.446 9

0118 Ethel NE 0.446 4 0110 Lori AZ 0.358 6 0112 Pat WY 0.328 1 0102 Will

MA 0.314 3 0105 John AZ 0.265 2 0104 Sue NY 0.159 } } } high medium

low 124 Layers of Data Abstraction SEMMA starts with data There are many different levels of data within an organization Think of a pyramid The most abundant source is operational data every transaction, bill, payment, etc. at bottom of pyramid Business rules tell us what weve learned from the data at top of pyramid Other layers in between 125 SEMMA: Select and Sample What data is available? Where does it come from? How often is it updated? When is it available? How recent is it? Is internal data sufficient? How much history is needed? 126 Data Mining Prefers Customer Signatures Often, the data come from many different sources Relational database technology allows us to construct a customer signature from these multiple sources

The customer signature includes all the columns that describe a particular customer the primary key is a customer id the target columns contain the data we want to know more about (e.g., predict) the other columns are input columns 127 Profiling is a Powerful Tool Profiling involves finding patterns from the past and assuming they will remain valid The most common approach is via surveys Surveys tell us what our customers and prospects look like Typical profiling question: What do churners look like? Profiling is frequently based on demographic variables e.g., location, gender, age 128 Profiling has its Limitations Even at its best, profiling tells us about the past Connection between cause and effect is sometimes unclear people with brokerage accounts have a minimal balance in their savings account customers who churn are those who have not used their telephones (credit cards) for the past month customers who use voicemail make a lot of short calls to the same number

More appropriate for advertising than one-to-one marketing 129 Two Ways to Aim for the Target Profiling: What do churners look like? data in input columns can be from the same time period (the past) as the target Prediction: Build a model that predicts who will churn next month data from input columns must happen before the target data comes from the past the present is when new data are scored 130 The Past Needs to Mimic the Present Past Distant Past ends here Recent Past starts here Present Data ends here Future

Predictions start here We mimic the present by using the distant past to predict the recent past 131 How Data from Different Time Periods are Used Jan Feb Model Set 4 Mar Apr May 3 Score Set 2 1 4 3 Jun Jul Aug Sep

+1 2 1 +1 The model set is used to build the model The score set is used to make predictions It is now August X marks the month of latency Numbers to left of X are months in the past 132 Multiple Time Windows Help the Models Do Well in Predicting the Future Model Set Jan Feb 4 Score Set Mar Apr May 3 2 1

4 3 2 1 4 3 Jun Jul Aug Sep +1 +1 2 1 +1 Multiple time windows capture a wider variety of past behavior They prevent us from memorizing a particular season 133 Rules for Building a Model Set for a Prediction

All input columns must come strictly before the target There should be a period of latency corresponding to the time needed to gather the data The model set should contain multiple time windows of data 134 More about the Model and Score Sets The model set can be partitioned into three subsets the model is trained using pre-classified data called the training set the model is refined, in order to prevent memorization, using the test set the performance of models can be compared using a third subset called the evaluation or validation set The model is applied to the score set to predict the (unknown) future 135 Stability Challenge: Memorizing the Training Set Error Rate Training Data Model Complexity Decision trees and neural networks can memorize nearly any pattern in the training set

136 Danger: Overfitting This is the model we want Error Rate The model has overfit the training data Te st Tra i nin Da ta gD at a As model complexity grows, performance deteriorates on test data Model Complexity 137

Building the Model from Data Both the training set and the test set are used to create the model Algorithms find all the patterns in the training set some patterns are global (should be true on unseen data) some patterns are local (only found in the training set) We use the test set to distinguish between the global patterns and the local patterns Finally, the validation set is needed to evaluate the models performance 138 SEMMA: Explore the Data Look at the range and distribution of all the variables Identify outliers and most common values Use histograms, scatter plots, and subsets Use algorithms such as clustering and market basket analysis Clementine does some of this for you when you load the data 139 SEMMA: Modify Add derived variables total, percentages, normalized ranges, and so on extract features from strings and codes Add derived summary variables median income in ZIP code Remove unique, highly skewed, and correlated

variables often replacing them with derived variables Modify the model set 140 The Density Problem The model set contains a target variable fraud vs. not fraud churn vs. still a customer Often binary, but not always The density is the proportion of records with the given property (often quite low) fraud 1% churn 5% Predicting the common outcome is accurate, but not helpful 141 Back to Oversampling 1 2 3 4 5 6

7 8 9 1 0 2 6 1 0 1 1 1 2 1 3 1 4 1 5 1

6 1 7 1 8 1 9 2 0 1 2 1 7 2 0 2 1 2 2 2 3 2

4 2 5 2 6 2 7 2 8 2 9 3 0 2 3 2 9 3 0 3 1 3

2 3 3 3 4 3 5 3 6 3 7 3 7 3 9 4 0 3 1 3 4 4

0 4 1 4 2 4 3 4 4 4 5 4 6 4 7 4 8 4 9 5 0 4

5 4 8 5 0 Original data has 45 white and 5 dark (10% density) The model set has 10 white and 5 dark (33% density ) For every 9 white (majority) records in the original data, two are in the oversampled model set Oversampling rate is 9/2 = 4.5 142 Two Approaches to Oversampling Build a new model set of the desired density fewer rows takes less time to build models more time for experimentation in practice, aim for at least 10,000 rows Use frequencies to reduce the importance of some rows uses all of the data Use a density of approx. 50% for binary outcomes 143 Oversampling by Taking a Subset of the Model Set

ID Name State Flag 1 0102 Will MA F 2 0104 Sue NY 3 0105 John 4

0110 5 ID Name State Flag 1 0102 Will MA F F 3 0105 John AZ

F AZ F 5 0111 Beth NM T Lori AZ F 6 0112 Pat WY F 0111

Beth NM T 8 0117 Frank MS T 6 0112 Pat WY F 9 0118 Ethel

NE F 7 0116 David ID F 8 0117 Frank MS T 9 0118 Ethel NE F

The original data has 2 Ts and 7 Fs (22% density) Take all the Ts and 4 of the Fs (33% density) The oversampling rate is 7/4 = 1.75 144 Oversampling via Frequencies ID Name State Flag ID Name State Flag Frq 1 0102 Will MA

F 1 0102 Will MA F 0.5 2 0104 Sue NY F 2 0104 Sue NY

F 0.5 3 0105 John AZ F 3 0105 John AZ F 0.5 4 0110 Lori AZ

F 4 0110 Lori AZ F 0.5 5 0111 Beth NM T 5 0111 Beth NM

T 1.0 6 0112 Pat WY F 6 0112 Pat WY F 0.5 7 0116 David ID

F 7 0116 David ID F 0.5 8 0117 Frank MS T 8 0117 Frank MS T

1.0 9 0118 Ethel NE F 9 0118 Ethel NE F 0.5 Add a frequency or weight column for each F, Frq = 0.5 for each T, Frq = 1.0 The model set has density of 2/(2 + 0.5 7) = 36.4% The oversampling rate is 7/3.5 = 2 145 SEMMA: Model Choose an appropriate technique

decision trees neural networks regression combination of above Set parameters Combine models 146 Regression Tries to fit data points to a known curve (often a straight line) Standard (well-understood) statistical technique Not a universal approximator (form of the regression needs to be specified in advance) 147 Neural Networks Based loosely on computer models of how brains work Consist of neurons (nodes) and arcs, linked together Each neuron applies a nonlinear function to its inputs to produce an output Particularly good at producing numeric outputs No explanation of result is provided 148 Decision Trees Looks like a game of Twenty Questions At each node, we fork based on variables e.g., is household income less than $40,000?

These nodes and forks form a tree Decision trees are useful for classification problems especially with two outcomes Decision trees explain their result the most important variables are revealed 149 Experiment to Find the Best Model for Your Data Try different modeling techniques Try oversampling at different rates Tweak the parameters Add derived variables Remember to focus on the business problem 150 It is Often Worthwhile to Combine the Results from Multiple Models ID Name State Mod 1 Mod 2 Mod 3

1 0102 Will MA 0.111 0.314 0.925 2 0104 Sue NY 0.121 0.159 0.491 3 0105

John AZ 0.133 0.265 0.211 4 0110 Lori AZ 0.146 0.358 0.692 5 0111 Beth NM 0.411

0.979 0.893 6 0112 Pat WY 0.510 0.323 0.615 7 0116 David ID 0.105 0.879 0.298

8 0117 Frank MS 0.116 0.502 0.419 9 0118 Ethel NE 0.152 0.446 0.611 151 Multiple-Model Voting Multiple models are built using the same input data Then a vote, often a simple majority or plurality

rules vote, is used for the final classification Requires that models be compatible Tends to be robust and can return better results 152 Segmented Input Models Segment the input data by customer segment by recency Build a separate model for each segment Requires that model results be compatible Allows different models to focus and different models to use richer data 153 Combining Models What is response to a mailing from a non-profit raising money (1998 data set) Exploring the data revealed the more often, the less money one contributes each time so, best customers are not always most frequent Thus, two models were developed who will respond? how much will they give? 154 Compatible Model Results In general, the score refers to a probability for decision trees, the score may be the actual density

of a leaf node for a neural network, the score may be interpreted as the probability of an outcome However, the probability depends on the density of the model set The density of the model set depends on the oversampling rate 155 An Example The original data has 10% density The model set has 33% density Each white in model set represents 4.5 white in original data Each dark represents one dark The oversampling rate is 4.5 156 A Score Represents a Portion of the Model Set Suppose an algorithm identifies the group at right as most likely to churn The score would be 4/6 = 67%, versus the density of 33% for the entire model set This score represents the probability on the oversampled data This group has a lift of 67/33 = 2 157 Determining the Score on the Original

Data The corresponding group in the original data has 4 dark and 9 white, for a score of 4 / (4 + 9) = 30.7% The original data has a density of 10% The lift is now 30.7/10 = 3.07 158 Determining the Score -- continued The original group accounted for 6/15 = 40% of the model set In the original data, it corresponds to 13/50 = 26% Bottom line: before comparing the scores that different models produce, make sure that these scores are adjusted for the oversampling rate The final part of the SEMMA process is to assess the results 159 Confusion Matrix (or Correct Classification Matrix) Predicted Actual Yes No Yes 800

50 No 50 100 There are 1000 records in the model set When the model predicts Yes, it is right 800/850 = 94% of the time When the model predicts No, it is right 100/150 = 67% of the time The density of the model set is 150/1000 = 15% 160 Confusion Matrix-- continued The model is correct 800 times in predicting Yes The model is correct 100 times in predicting No The model is wrong 100 times in total The overall prediction accuracy is 900/1000 = 90% 161 From Data to the Confusion Matrix Name State

Score Act Pred 1 0102 Will MA 0.314 F F 2 0104 Sue NY 0.159 F F 3 0105 John AZ 0.265

F F 4 0110 Lori AZ 0.358 T F 5 0111 Beth NM 0.979 T T 6 0112 Pat WY 0.328 F F

7 0116 David ID 0.446 F T 8 0117 Frank MS 0.897 T T 9 0118 Ethel NE 0.446 F T Actual

Predicted ID T F T 2 2 F 1 4 We hold back a portion of the data so we have scores and actual values The top tercile is given a predicted value of T Because of tie, we have 4 Ts predicted 162 How Oversampling Affects the Results Actual Yes No

Yes 800 50 No 50 100 Predicted Predicted Actual Yes Yes No 8000 50 500 100 No

Model set Original data The model set has a density of 15% No Suppose we achieve this density with an oversampling rate of 10 So, for every Yes in the model set there are 10 Yess in the original data 163 How Oversampling Affects the Results-continued Original data has a density of 150/8650 = 1.734% We expect the model to predict No correctly 100/600 = 16.7% of the time The accuracy has gone down from 67% to 16.7% The results will vary based upon the degree of oversampling 164 Lift Measures How Well the Model is Doing The density of dark in the model set is 33.3% The density of dark in the subset chosen by the model is 66.7% The lift is 66.7/33.3 = 2 The model is doing twice as well as choosing circles at random 165 Lift on a Small Data Set

ID Name State Score Act Pred Tercile 1 0102 Will MA 0.314 F F 3 2 0104 Sue NY 0.159 F

F 3 3 0105 John AZ 0.265 F F 3 4 0110 Lori AZ 0.358 T F 2 5 0111 Beth NM

0.979 T T 1 6 0112 Pat WY 0.328 F F 2 7 0116 David ID 0.446 F T 1 8 0117 Frank

MS 0.897 T T 1 9 0118 Ethel NE 0.446 F T 2 Note: we break tie arbitrarily Model set density of T is 33.3% Tercile 1 has two T and one F Tercile 1 has a density of 66.7% The lift is 66.7/33.3 = 2 166

The Lift Chart for the Small Data Set 2.5 2.0 1.5 1.0 0.5 Tercile 1 has a lift of 66.7/33.3 = 2 Terciles 1 and 2 have a density of 3/6 = 50% and a lift of 50/33.3 = 1.5 Since terciles 1, 2, 3 1 2 and 3 comprise the Tercile entire model set, the lift is 1 We always look at lift in a cumulative sense 167 Cumulative Gains Chart * 100% * el

d o m h * m od * ra nd om 33% 33% Cumulative gains chart shows the proportion of responders (churners) in each tercile (decile) Horizontal axis shows the tercile (decile) el 67% t wi

* Vertical axis gives the proportion of responders that model yields 67% 100% The cumulative gains chart and the lift chart are related 168 More on Lift and the Cumulative Gains Chart The lift curve is the ratio of the cumulative gains of the model to the cumulative gains of the random model The cumulative gains chart has several advantages it always goes up and to the right can be used diagnostically Reporting the lift or cumulative gains on the training set is cheating Reporting the lift or cumulative gains without the oversampling rate is also cheating 169 Summary of Data Mining Process Cycle of Data Mining identify the right business problem transform data into actionable information act on the information measure the effect

SEMMA select/sample data to create the model set explore the data modify data as necessary model to produce results assess effectiveness of the models 170 The Data in Data Mining Data comes in many forms internal and external sources Different sources of data have different peculiarities Data mining algorithms rely on a customer signature one row per customer multiple columns See Chapter 6 in Mastering Data Mining for details 171 Preventing Customer Attrition We use the noun churn as a synonym for attrition We use the verb churn as a synonym for leave Why study attrition? it is a well-defined problem it has a clear business value we know our customers and which ones are valuable we can rely on internal data the problem is well-suited to predictive modeling

172 When You Know Who is Likely to Leave, You Can Focus on keeping high-value customers Focus on keeping high-potential customers Allow low-potential customers to leave, especially if they are costing money Dont intervene in every case Topic should be called managing customer attrition 173 The Challenge of Defining Customer Value We know how customers behaved in the past We know how similar customers behaved in the past Customers have control over revenues, but much less control over costs we may prefer to focus on net revenue rather than profit We can use all of this information to estimate customer worth These estimates make sense for the near future 174 Why Maturing Industries Care About Customer Attrition Representative Growth in a Maturing Market Number of Customers (1000s) 1200 total customers

As growth flattens out, customer attrition becomes more important 1000 800 600 In this region of rapid growth, building infrastructure is more important than CRM 400 new customers 200 churners 0 0 1 2 3 4 5

6 Year 7 8 9 10 11 175 Number of Churners per 100 New Customers Another View of Customer Attrition In a fast growth market, almost all customers are new customers 100 90 80 70 60 50 40 30 20

10 0 1 2 3 4 5 6 7 8 Time Period 9 10 11 12 13 14 15

In a maturing market, customer attrition and customer profitability become issues In a mature market, the amount of attrition almost equals the number of new customers 176 Why Attrition is Important When markets are growing rapidly, attrition is usually not important customer acquisition is more important Eventually, every new customer is simply replacing one who left Before this point, it is cheaper to prevent attrition than to spend money on customer acquisition One reason is that, as a market becomes saturated, acquisition costs go up 177 In Maturing Markets, Acquisition Response Rates Decrease and Costs Increase Cost per Response $250 $200 $150

220 Assumption: $1 per contact, $20 offer 120 $100 70 53.3 $50 45 40 $0 0.00% 1.00% 2.00% 3.00% 4.00% 5.00% 6.00% Response Rate

178 Acquisition versus Retention Cost per Response $250 $200 $150 220 Assumption: $1 per contact, $20 offer 120 $100 70 53.3 $50 45 40 $0 0.00% 1.00% 2.00%

3.00% 4.00% 5.00% 6.00% Response Rate 179 Acquisition versus Retention-- continued As response rate drops, suppose we spend $140 to obtain a new customer Alternatively, we could spend $140 to retain an existing customer Assume the two customers have the same potential value Some possible options decrease the costs of the acquisition campaign implement a customer retention campaign combination of both 180 Retention and Acquisition are Different Economics of acquisition campaigns customers are unlikely to purchase unless contacted/invited cost of acquisition is the campaign cost divided by the number acquired during the campaign Economics of retention campaigns

customers are targeted for retention, but some would have remained customers anyway cost of retention is the campaign cost divided by the net increase in retained customers 181 Cost per Retention: How Many Responders Would Have Left? For a retention campaign, we need to distinguish between customers who merely respond and those who respond and would have left If the overall group has an attrition rate of 25%, you can assume that one quarter of the responders are saved Since people like to receive something for nothing, response rates tend to be high As before, assume $1 per contact and $20 per response We need to specify response rate and attrition rate 182 Cost Per Retention by Attrition Rate $1,600 5.00% $1,400 10.00% $1,200 15.00% 20.00%

Cost Per Retention $1,000 25.00% 35.00% $800 $600 $400 $200 $0 0.0% 2.0% 4.0% 6.0% 8.0% 10.0% 12.0% 14.0%

16.0% 18.0% 20.0% 22.0% Response Rate of Compaign 183 A Sample Calculation Given: $1 per contact, $20 per response What is the cost per retention with response rate of 20% and attrition rate of 10%? Suppose 100 people are contacted 20 people respond 20 10% = 2 would have left, but are retained campaign cost = $100 + $20 20 = $500 cost per retention = $500/2 = $250 184 Typical Customer Retention Data Num Customers Beg of Year Num Churners New Customers Num Customers End of Year

40% 70% 0 0 100,000 1990 1991 100,000 130,000 40,000 52,000 70,000 91,000 130,000 169,000 1992 169,000 67,600 118,300

219,700 1993 219,700 87,880 153,790 285,610 1994 285,610 114,244 199,927 371,293 1995 371,293 148,517 259,905 482,681

1996 482,681 193,072 337,877 627,485 1997 627,485 250,994 439,240 815,731 1998 815,731 326,292 571,012 1,060,450 Each year, 40% of existing customers leave Each year, 70% new customers are added At year end, the number of customers is the number at the end of the

previous year minus the number who left plus the number of new customers 185 How to Lie with Statistics Num Customers Beg of Year Num Churners New Customers Num Customers End of Year Reported Churn 40% 70% 0 0 100,000 0 1990 1991

100,000 130,000 40,000 52,000 70,000 91,000 130,000 169,000 30.77% 30.77% 1992 169,000 67,600 118,300 219,700 30.77% 1993 219,700 87,880

153,790 285,610 30.77% 1994 285,610 114,244 199,927 371,293 30.77% 1995 371,293 148,517 259,905 482,681 30.77% 1996

482,681 193,072 337,877 627,485 30.77% 1997 627,485 250,994 439,240 815,731 30.77% 1998 815,731 326,292 571,012 1,060,450 30.77%

When we divide by end-of-year customers, we reduce the attrition rate to about 31% per year This may look better, but it is cheating 186 Suppose Acquisition Suddenly Stops Num Customers Beg of Year Num Churners New Customers Num Customers End of Year Reported Churn 40% 70% 0 0 100,000 0

1990 1991 100,000 130,000 40,000 52,000 70,000 91,000 130,000 169,000 30.77% 30.77% 1992 169,000 67,600 118,300 219,700 30.77% 1993 219,700

87,880 153,790 285,610 30.77% 1994 285,610 114,244 199,927 371,293 30.77% 1995 371,293 148,517 259,905 482,681 30.77%

1996 482,681 193,072 337,877 627,485 30.77% 1997 627,485 250,994 439,240 815,731 30.77% 1998 1999 815,731 1,060,450 326,292 424,180

571,012 0 1,060,450 636,270 30.77% 66.67% If acquisition of new customers stops, existing customers still leave at same rate But, churn rate more than doubles 187 Measuring Attrition the Right Way is Difficult The right way gives a value of 40% instead of 30.77% who wants to increase their attrition rate? this happens when the number of customers is increasing For our small example, we assume that new customers do not leave during their first year the real world is more complicated we might look at the number of customers on a monthly basis 188 Effect of Attrition on Revenue Effect of Attrition on Revenue

$25,000,000 N e w C u s t o me r R e v e n u e $20,000,000 Revenue E x i s t i n g C u s t o me r R e v e n u e $15,000,000 L o ss D u e t o A t t r i t i o n $10,000,000 $5,000,000 $0 -$5,000,000 90 91 92 93 Time 94 95 96

97 98 Assumption: $20 for existing customers; $10 for new customers 189 Effect of Attrition on Revenues -continued From previous page, the loss due to attrition is about the same as the new customers revenue The loss due to attrition has a cumulative impact If we could retain some of these customers each year, they would generate revenue well into the future It is useful to examine the relationship between attrition and the length of the customer relationship Often, attrition is greater for longer-duration customers 190 Relationship Between Attrition and the Length of the Customer Relationship 500,000 450,000 400,000 Current Customers Churners Num Customers

350,000 300,000 250,000 200,000 150,000 100,000 50,000 0 0 1 2 3 4 5 6 7 8 Length of Relationship 191 What is a Customer Attrition Score? When building an attrition model, we seek a score for each customer This adds a new column to the data

Two common approaches a relative score or ranking of who is going to leave an estimate of the likelihood of leaving in the next time period 192 Attrition Scores are an Important Part of the Solution Having an idea about which customers will leave does not address key business issues why are they leaving? where are they going? is brand strength weakening? are the products still competitive? However, it does allow for more targeted marketing to customers it is often possible to gain understanding and combat attrition while using attrition models 193 Requirements of Effective Attrition Management Keeping track of the attrition rate over time Understanding how different methods of acquisition impact attrition Looking at some measure of customer value, to determine which customers to let leave Implementing attrition retention efforts for high-value customers Knowing who might leave in the near future 194

Three Types of Attrition Voluntary attrition when the customer goes to a competitor our primary focus is on voluntary attrition models Forced attrition when the company decides that it no longer wants the customer and cancels the account often due to non-payment our secondary focus is on forced attrition models Expected attrition when the customer changes lifestyle and no longer needs your product or service 195 What is Attrition? In the telecommunications industry it is very clear customers pay each month for service customers must explicitly cancel service the customer relationship is primarily built around a single product It is not so clear in other industries retail banking credit cards retail e-commerce 196 Consider Retail Banking Customers may have a variety of accounts

deposit and checking accounts savings and investment mortgages loans credit cards business accounts What defines attrition? closing one account? closing all accounts? something else? 197 Retail Banking--Continued One large bank uses checking account balance to define churn What if a customer (e.g., credit card or mortgage customer) does not have a checking account with the bank? Another example from retail banking involves forced attrition Consider the bank loan timeline on the next page 198 The Path to Forced Attrition Good Customer Missed Payment 30 Days Late 60 Days Late Forced Attrition: the bank sells the loan to a collection

agency 90 Days Late 199 Credit Cards Credit cards, in the U.S., typically dont have an annual fee there is no incentive for a cardholder to tell you when she no longer plans on using the card Cardholders typically use the card several times each month Customers who stop using the card and are not carrying a balance are considered to be silent churners The definition of silent churn varies among issuers customers who have not used the card for six months customers who have not used the card for three months and for nine of the last twelve months 200 The Catalog Industry Like credit cards, catalogs are free therefore, customers have little incentive to cancel them Unlike credit cards, purchases from catalogs are more sporadic so defining silent churn is more difficult Purchases from catalogs are often seasonal so silent churn is measured over the course of years, rather than months

201 E-commerce Attrition in e-commerce is hardest of all to define consider Amazon Web sites are free, so there is no incentive to announce intention to churn Customers often visit many times before making a purchase Some customers buy frequently and others dont how can Amazon decrease time between purchases? The e-commerce world is still growing rapidly, so customer retention is not a major issue But it will become a major issue soon 202 Ways to Address Voluntary Attrition Allow customers who are not valuable to leave Offer incentives to stay around for a period of time teaser rates on credit cards, free weekend airtime, no payments or interest for six months Offer incentives to valuable customers discounts/extras Stimulate usage miles for minutes, donations to charities, discounts 203 Predicting Voluntary Attrition

Can Be Dangerous Voluntary attrition sometimes looks like expected or forced attrition Confusing voluntary and expected attrition results in the waste of marketing dollars Confusing voluntary and forced attrition means you lose twice by, again, wasting marketing dollars by incurring customer non-payment 204 Ways to Address Forced Attrition Stop marketing to the customer no more catalogs, billing inserts, or other usage stimulation Reduce credit lines Increase minimum payments Accelerate cancellation timeline Increase interest rates to increase customer value while he is still paying 205 Predicting Forced Attrition Can Be Dangerous Forced attrition sometimes looks at good customers they have high balances they are habitually late Confusing forced attrition with good customers may encourage them to leave Confusing forced with voluntary attrition may

hamper winback prospects 206 Attrition Scores are Designed for the Near Future The chance that someone will leave tomorrow is essentially 0% The chance that someone will leave in the next 100 years is essentially 100% A good attrition score is valid for one to three months in the future Attrition modeling can take place at any point in the customer lifecycle 207 A Related Problem: Estimating the Customer Lifetime We also want to estimate the length of a customers lifetime If we know how long someone will remain a customer, we know when he will churn This can help us obtain an estimate of customer profitability Customers with a high lifetime value (profitability) are the ones we want to prioritize for retention efforts 208 Suppose Customers Have a 10% Chance of Leaving in the Next Month Probability the customer leaves in next month = .1 10% of customers have a lifetime of 1 month (round up)

Probability the customer leaves in 2 months = .9 .1= .09 9% have a lifetime of 2 months (round up) Probability the customer leaves in 3 months = .9 .9 .1= .081 8.1% have a lifetime of 3 months (round up) Probability that customer leaves in x months = (.9)x1 .1 209 Average Customer Lifetime In statistics, we refer to this as a geometric distribution If the churn rate per month is x%, the average lifetime is 1/x% months In this example, the average lifetime is 1/.1= 10 months An ideal attrition model says all customers leaving in the next month get a score of 1 or 100% all other customers get a score of 0 or 0% 210 Attrition Modeling Summary It is very important to manage attrition, especially in maturing industries Slowing attrition may be the cheapest way to maintain a critical mass of customers There are different types of attrition voluntary, forced, expected Related to attrition is the customer lifetime useful in calculating lifetime customer value 211

CRM Sources The vast majority of these CRM slides has been borrowed, adapted, or reworked from one of the two sources below 1. Michael Berry and Gordon Linoff, Customer Relationship Management Through Data Mining, SAS Institute, 2000 Michael Berry and Gordon Linoff, Mastering Data Mining, John Wiley & Sons, 2000 2. I have also consulted Data Mining Techniques (Wiley, 1997) by Berry and Linoff, in preparing these slides 212 Decision Trees and Churn Modeling Learn about decision trees what are they? how are they built? advantages versus disadvantages Case study from the cellular telephone industry problem definition looking at the results other techniques for modeling churn 213 Data Mining with Decision Trees

A decision tree is a set of rules represented in a tree structure Easy to understand how predictions are made Easy to build and visualize Can be used for binary or multiple outcomes Roots, nodes, leaves, and splits 214 Data Mining with Decision Trees--continued We build the decision tree using the training and test sets The decision tree allows us to make predictions understand why certain predictions make sense understand which variables are most important spot unexpected patterns There is one path from the root to any leaf A given data record is associated with a single leaf 215 A Path Through the Decision Tree At each node, a decision is made--which variable to split These variables are the most important Each leaf should be as pure as possible (e.g., nearly all

churns) All records landing at the same leaf get the same prediction A small example follows 216 A Decision Tree for Widget Buyers 1 yes 6 no Res NY 11 yes 9 no No 2 yes 3 no Age > 35 Res = NY 10 yes 3 no Age <= 35 8 yes Rule 1. If residence not NY, then not a widget buyer Rule 2. If residence NY and age > 35, then not a widget buyer Yes

# correctly classified Accuracy = Rule 3. If residence NY and age <= 35, then a widget buyer Adapted from (Dhar & Stein, 1997) No = total # 6 38 0.85 20 217 Building a Decision Tree We start at the root node with all records in the training set Consider every split on every variable Choose the one that maximizes a measure of purity ideally, all churners on left and non-churners on right For each child of the root node, we again search for the best split i.e., we seek to maximize purity Eventually, the process stops no good split available or leaves are pure 218

Building a Decision Tree--continued The above process is sometimes called recursive partitioning To avoid overfitting, we prune the tree using the test set to prune is to simplify After the tree has been built, each leaf node has a score A leaf score may be the likelihood that the more common class arises in training and test sets 219 The Leaves Contain the Scores Overall, the most common class is No 11,112 records (including Sams) are associated with this leaf Yes 96.5% are not fraudulent 3.5% No 96.5% Size

11,112 Sam gets a score of .965 220 Scoring Using a Decision Tree ID Name State Score 1 0102 Will MA 0.428 2 0112 Pat WY 0.283 3 0116 David

ID 0.957 4 0117 Frank MS 0.892 5 0118 Ethel NE 0.957 This example shows the relationship between records in a table and decision tree leaves Remember, each record will fall in exactly one leaf 221 A Real Life Example: Predicting Churn Training Test 13.5% 86.5% 39,628

13.8% 86.2% 19,814 Churn data from the cellular industry Begin at the root node The training set is twice as large as the test set Both have approximately the same percentage of churners (13.5%) The tree will be built using training set data only222 The First Split 13.5% 86.5% 39,628 < 0.7% 13.8% 86.2% 19,812 Handset Churn Rate Training set on left, test set on right 3.8% < 3.8% 3.5% 3.0% 96.5% 97.0% 11,112 5,678

14.9% 85.1% 23,361 15.6% 84.4% 11,529 28.7% 71.3% 5,155 29.3% 70.7% 2,607 223 The First Split--continued The first split is made on the variable Handset Churn Rate Handsets drive churn in the cellular industry So, first split is no surprise The algorithm splits the root node into three groups low, medium, and high churn rates If we look at the child on the far right, we see that the churn rate has more than doubled Far-right child has a lift of 29.3/13.8 = 2.12 224 As the Tree Grows Bigger, Nodes and Leaves Are Added

13.5% 86.5% 39,628 < 0.7% 3.5% 3.0% 96.5% 97.0% 11,112 5,678 < 4855 67.3% 66.0% 32.7% 34.0% 110 47 13.8% 86.2% 19,814 Handset Churn Rate < 3.8% 14.9% 15.6% 85.1% 84.4% 23,361 11,529 < 0.0056 58.0% 56.6% 42.0% 43.4% 219 99 Total Amt Overdue < 88455 70.9% 52.0%

29.1% 48.0% 55 25 3.8% 28.7% 71.3% 5,155 < 0.18 39.2% 40.4% 60.8% 59.6% 148 57 29.3% 70.7% 2,607 CALL0/CALL3 >= 0.18 27.0% 27.9% 73.0% 72.1% 440 218 >= 88455 25.9% 44.4% 74.1% 55.6% 54 27 225

As the Tree Grows--continued CALL0 = calls in the most recent month CALL3 = calls four months ago CALL0/ CALL3 is a derived variable The number of calls has been decreasing Total Amt Overdue = total amount of money overdue on account When a large amount of money is due, the voluntary churn rate goes down It is easy to turn the decision tree into rules 226 Three of the Best Rules If Handset Churn Rate 3.8% AND CALL0/CALL3 < 0.0056 AND Total Amt Overdue < 88455 THEN churn likelihood is 52.0% (on test set) If Handset Churn Rate 3.8% AND CALL0/CALL3 < 0.0056 AND Total Amt Overdue < 4855 THEN churn likelihood is 66.0% (on test set) If Handset Churn Rate 3.8% AND CALL0/CALL3 < 0.18 THEN churn likelihood is 40.4% (on test set) 227 From Trees to Boxes Another useful visualization tool is a box chart

The lines in the box chart represent specific rules The size of the box corresponds to the amount of data at a leaf The darkness of a box can indicate the number of churners at a leaf (not used here) We use HCR and C03 as abbreviations for Handset Churn Rate and CALL0/CALL3 Fill in three boxes in top right on page 228 228 From Trees to Boxes--continued HCR HCR < 0.7% 0.7% HCR < 3.8% 3.8% C03 < 0.18 C03 0.0056 HCR C03 3.8%

0.18% 229 Finding a Good Split at a Decision Tree Node There are many ways to find a good split But, they have two things in common splits are preferred where the children are similar in size splits are preferred where each child is as pure as possible Most algorithms seek to maximize the purity of each of the children This can be expressed mathematically, but it is not our focus here 230 Synonymous Splits There are many ways to determine that the number of calls is decreasing low amount paid for calls small number of local calls small amount paid for international calls Different variables may be highly correlated or essentially synonymous Decision trees choose one of a set of synonymous variables to split the data These variables may become input variables to neural network models 231

Lurking Inside Big Complex Decision Trees Are Simpler, Better Ones 232 Problem: Overfitting the Data 13.5% 13.8% 86.5% 86.2% 39,628 19,814 Handset Churn Rate 3.8% Good! The training set and test set are about the same 28.7% 71.3% 5,155 Good! The training set and test set are about the same 29.3% 70.7% 2,607 CALL0/CALL3 < 0.0056

58.0% 42.0% 219 Total 70.9% 29.1% 55 56.6% 43.4% 99 Amt Overdue < 88455 52.0% 48.0% 25 Good! The training set and test set are about the same X Ouch! The tree has memorized the training set, but not generalized 233 Sometimes Overfitting is Evident by Looking at the Accuracy of the Tree Proportion Correctly Classified 0.76 0.75 TRAIN 0.74

TEST 0.73 0.72 0.71 0.70 0.69 0 5 10 15 20 25 30 35 40 45 50 55 60

65 70 Number of Leaves 234 Pruning the Tree Prune here This is a good model Error rate is minimized on unseen data Error Rate Te st d a ta Tra inin gd ata Number of Leaves 235

Connecting Decision Trees and Lift A decision tree relating to fraud within the insurance industry is presented next Assume the analysis is based on test set data The cumulative gains chart consists of a series of line segments Each segment corresponds to a leaf of the tree The slope of each line corresponds to the lift at that leaf The length of each segment corresponds to the number of records at the leaf 236 Relationship Between a Cumulative Gains Chart and a Decision Tree No (3692) 80.0% Yes (923) 20.0% Base Policy All Perils (1481) Collision (1840) No (1029) 69.5% No (1405) 76.4% Yes (452) 30.5% Yes (435) 23.6% Fault Policy Holder (1042) No (606) 58.2% Yes (436) 41.8% Liability (1294) No (1258) 97.2% Yes (36) 2.8% Third Party(439)

No (423) 96.4% Yes (16) 3.6% 100.0% 90.0% 80.0% Rural (122) Urban (920) No (56) 45.9% Yes (66) 54.1% No (550) 59.8% Yes (370) 40.2% Adapted from Berry & Linoff (2000) %Fraud Accident Area 70.0% 60.0% 50.0% 40.0% Decision Tree 30.0% Baseline

20.0% 10.0% 0.0% 0.0% 10.0% 20.0% 30.0% 40.0% 50.0% 60.0% 70.0% 80.0% 90.0% 100.0% % Claims 237 Decision Trees: A Summary It is easy to understand results Can be applied to categorical and ordered inputs Finds inputs (factors) that have the biggest impact on the output

Works best for binary outputs Powerful software is available Care must be taken to avoid overfitting 238 Case Study: Churn Prediction Predicting who is likely to churn at a cellular phone company The cellular telephone industry is rapidly maturing The cost of acquiring new customers is rather high A handset is discounted in exchange for a time commitment As the market matures, it becomes cheaper to retain than acquire customers Review the difference between voluntary, forced, and expected attrition 239 The Problem This case study took place in a foreign country Outside of North America and Europe Mobil telephones are rather advanced in newly industrialized countries used instead of fixed line technology The penetration of mobile phones in the U.S. was 25% in 1999 240 The Problem-- continued The penetration of mobile phones in Finland was >50% in 1999 The company is the dominant cellular carrier in

its market It has 5 million customers Four or five competitors each have about 2 million customers 241 The Problem-- continued The company has a churn rate of about 1% per month It expects its churn rate to jump There are companies in this industry with churn rates of 4% per month The company wants to assign a propensity-to-churn score to each customer and put the resulting model into production ASAP They decided to develop the expertise in-house, using consultants The work took place over a two month period 242 Customer Base Segment # Customers % of Customers Churn Rate Elite 1,500,000

30 1.3% Non Elite 3,500,000 70 0.9% Total 5,000,000 100 1.1% Elite customers exceeded some threshold of spending in the previous year They have been around at least one year Why is their churn rate higher? 243 Data Inputs Customer information file telephone number, billing plan, zip code, additional services, Elite Club grade, service activation date, etc. Handset information type, manufacturer, weight, etc. Dealer information marketing region and size

Billing history Historical churn rate information by demographics and handset type 244 The Model Set Uses Overlapping Time Windows Jan Feb Mar Apr May Model Set Model Set Score Set 6 Jun Jul Aug 5 4 3 2 1 6 5 4

3 2 1 6 5 4 3 Sep Oct Nov +1 +1 2 1 +1 For the model set, August and September churners are used We want to use the model to make predictions for November

245 Solving the Right Business Problem Initial problem: assign a churn score to all customers Complicating issues new customers have little call history telephones? individuals? families? voluntary churn versus involuntary churn how will the results be used? Revised problem: by Sept. 20th, provide a list of the 10,000 Elite customers who are most likely to churn in October The revised problem invites action 246 Build a Number of Churn Models Numerous decision tree models were built It was discovered that some customers were taking advantage of the family plan to obtain better handsets at minimal cost These customers used the marketing campaigns to their own advantage Different parameters were used to build different models The density of churners in the model set was one important parameter (50% worked well) 247 Database Marketing and the Catalog Industry Catalogs are everywhere For many people, catalogs are the preferred

method of shopping Their resemblance to magazines is intentional For many people, the reaction is the same Catalogs are very old the first Sears catalog was published in 1888 Catalogs flourished over the next half century railroad, coast-to-coast mail service, mass-produced goods 248 Catalogs are Part of the Retail Business Catalog retailers know their customers as well as their merchandise They communicate one-to-one and not just through branding and mass market advertising Ability to customize catalogs Closely related to B2C Web retailing Many companies (e.g., Eddie Bauer) have stores, catalogs, and a retail Web site 249 The Catalog Industry $95 billion industry in U.S. (Catalog Age, 1998) Catalog sales are approximately 10% of all U.S. retail sales B2C Web retail sales are approximately 1% of U.S. retail sales (1999) Most catalogs now have an online presence

It is becoming difficult to distinguish between Web retailers and catalog companies 250 What do Catalogers Care About? Merchandising how to display and promote the sale of goods and services? Layout where to place items in the catalog? which models to use? Response closer to the subject of this course personalization, customization, recommendation which of several pre-assembled catalogs should be sent to a specific customer based on an estimate of his response to each? 251 Example of a Big Catalog: Fingerhut Manny Fingerhut started database marketing in 1948 in Minneapolis sold seat covers on credit to new car buyers Fingerhut and specialty catalogs emerged over time As of the late 1990s $2 billion in annual sales 400 million catalogs/year (> 1 million/day) terabytes of transaction history customer and prospect file with 1,400 fields describing 30 million households

252 Different Catalogers Have Different Approaches Fingerhut aims for a less affluent audience 4 easy payments of $19.95 is a way to hide higher interest rates the most profitable customers dont have to be the wealthiest Victorias Secret sees its catalog as part of its branding every mailbox should have one Eddie Bauer has a multi-channel strategy catalog, retail stores, and Web are equal partners Many are aimed at a very narrow audience 253 The Luxury of Pooled Data Unlike other industries, the catalog industry promotes the sharing of data through a company called Abacus U.S. Air doesnt know about my travels on TWA or Northwest Airlines But a catalog company (e.g., Eddie Bauer) can decide to send you a catalog because of a purchase you made from a competing catalog (e.g., L.L. Bean) As a consumer, you may have observed that a purchase from catalog A sometimes triggers the arrival of catalog B 254

Abacus and the World Wide Web Abacus is the catalog industry infomediary 1,100 member companies maintains a database of over 2 billion transactions includes the vast majority of U.S. catalog purchases sells summarized data to member companies details of individual transactions are not revealed facilitates industry-wide modeling DoubleClick a leader in Web advertising, now owns Abacus indicates the convergence of online and off-line retailing 255 Another Source of Data Household data vendors companies that compile data on every U.S. household Acxiom, First Data, Equifax, Claritas, Experian, R.L. Polk hundreds of variables (demographic, lifestyle, etc.) database marketers use this data to assemble mailing lists Available data internal data on customers industry-wide data from Abacus demographic data from household data vendors 256 What Some Big Catalog Companies are Doing with Data Mining Neural networks are being used to forecast call center staffing needs

New catalog creation what should go into it? which products should be grouped together? who should receive the catalog? Campaign optimization 257 Eddie Bauer Eddie Bauer is an interesting case They seek to maintain a unified view of the customer across three channels give the customer a unified view of Eddie Bauer They are integrating all channels for CRM 400 stores catalogs Web sales 258 Vermont Country Store (VCS) Eddie Bauer and other large companies use CRM extensively What about smaller companies? VCS is a successful family-owned business VCS is a $60 million catalog retailer (late 1990s) VCS first used CRM to improve the targeting of their catalog to increase the response rate They used SAS Enterprise Miner to achieve a dramatic return on investment 259

Early Company History Founded by V. Orton in 1945, after he returned from WWII As a child, he had helped his father and grandfather run a Vermont country store His first catalog went out in 1945 or 1946 36 items 12 pages mailed to 1000 friends and acquaintances contained articles and editorials had the feel of a magazine 260 The Next Generation Son Lyman Orton took over in the 1970s he remains as owner and chairman the CEO is not a family member Lyman focused on the catalog side of the business VCS grew by about 50% per year during the 1980s without data mining The catalog industry, as a whole, expanded rapidly during the 1980s 261 The Industry Matures As the catalog industry prospered, it attracted new entrants Catalogs began to fill every mailbox Response rates declined Growth rates declined

The cost of paper and postage increased rapidly By 1995, more than one-third of catalog firms were operating in the red VCS was still profitable, but concerned 262 Vermont Country Store Today They seek to sell merchandise that doesnt come back -- to people who do $60 million in annual sales 350 employees Use Enterprise Miner to build response (predictive) models VCS sells the notion of a simple and wholesome (old-fashioned) New England lifestyle 263 Business Problem: Find the Right Customers The VCS vision does not appeal to everyone Historical data on responders and non-responders is used to predict who will respond to the next catalog Focus on existing customers upside: we need to select a subset of a pre-determined set of customers downside: we are unable to identify outstanding new prospects The goal of response modeling is to find the right customers 264 Possible Objectives or Goals

Increase response rate Increase revenue Decrease mailing costs Increase profit Increase reactivation rate for dormant customers Increase order values Decrease product returns 265 RFM: Common Sense in Retailing Recency: customers who have recently made a purchase are likely to purchase again Frequency: customers who make frequent purchases are likely to purchase again Monetary: customers who have spent a lot of money in the past are likely to spend more money now Each one of these variables is positively correlated with response to an offer RFM combines all three variables 266 Where to Begin Any of the previous goals can be addressed via data mining The first step is to define the goal precisely The specific goal determines the target variable Build the customer signatures orders placed by customer over time items purchased by category over time customer information from internal sources use customer zip code to add external data purchase household and neighborhood level data 267

FEMALE MALE Under 5 years 9,246,000 9,663,000 5 to 9 years 9,723,000 10,190,000 10 to 14 years 9,591,000 10,070,000 15 to 19 years 9,644,000 10,201,000 20 to 24 years 8,925,000

9,263,000 25 to 29 years 9,134,000 9,025,000 30 to 34 years 9,906,000 9,715,000 35 to 39 years 11,321,000 11,202,000 40 to 44 years 11,277,000 11,095,000 45 to 49 years 9,965,000 9,616,000 50 to 54 years

8,593,000 8,137,000 55 to 59 years 6,800,000 6,231,000 60 to 64 years 5,573,000 4,993,000 65 to 69 years 5,115,000 4,341,000 70 to 74 years 4,900,000 3,872,000 75 to 79 years 4,300,000

3,085,000 80 to 84 years 3,021,000 1,836,000 85 to 89 years 1,800,000 864,000 90 to 94 years 851,000 313,000 95 to 99 years 276,000 79,000 51,000 11,000 100 years & over

RFM and Beyond RFM uses recency/frequency/ monetary cells to determine who receives a catalog On left is the U.S. population (as of Nov. 1999) by age and gender In this section, we compare RFM and more sophisticated approaches 268 FEMALE MALE Under 5 years 48.90% 51.10% 5 to 9 years 48.83% 51.17% 10 to 14 years 48.78%

51.22% 15 to 19 years 48.60% 51.40% 20 to 24 years 49.07% 50.93% 25 to 29 years 50.30% 49.70% 30 to 34 years 50.49% 49.51% 35 to 39 years 50.26% 49.74% 40 to 44 years

50.41% 49.59% 45 to 49 years 50.89% 49.11% 50 to 54 years 51.36% 48.64% 55 to 59 years 51.98% 48.02% 60 to 64 years 52.74% 47.26% 65 to 69 years 54.09%

45.91% 70 to 74 years 55.86% 44.14% 75 to 79 years 58.23% 41.77% 80 to 84 years 62.20% 37.80% 85 to 89 years 67.57% 32.43% 90 to 94 years 73.11% 26.89% 95 to 99 years

77.75% 22.25% 100 years & over 82.26% 17.74% The Cells of the Table Below are Interesting More boys than girls until age 25 More females than males after age 25 75 79 year old men fought in WWII 269 Cell-Based Approaches are Very Popular RFM is used often in the catalog industry Cell-based approaches are used to assess credit risk in the banking industry (e.g., CapitalOne) The insurance industry uses cell-based approaches to assess risk Market research also uses cell-based approaches, especially when the cells reflect demographics that can be used to predict purchasing behaviors 270

RFM is a Powerful General Methodology Divide the population into cells based on known variables e.g., age, sex, income, bushiness of mustache Measure item of interest (e.g., response) for each cell (e.g., via a test mailing list) In full-sized mailing, focus on the cells of interest through selection, pricing, advertising 271 RFM: The Baseline Method for the Catalog Industry RFM was developed in the 1980s, when the catalog industry was growing rapidly Proved to be very valuable, so it spread rapidly Became less effective in the 1990s at a time when costs (including paper costs) started rising rapidly Not a customer-centric approach, but important for comparison purposes 272 Sort List by Recency Sort customers from most recent purchase date to least recent arrangement is based entirely on rank Break the list into equal-sized sections called quantiles top, bottom high, medium, low

quartiles quintiles deciles This quantile becomes the R component of the RFM cell assignment 273 Sort List by Frequency Resort the list by frequency, break the list into quantiles But how do we measure frequency? total number of orders divided by the number of months since the first order average number of orders per month over the past year The quantile becomes the F component of the RFM cell assignment 274 Sort List by Monetary Value Resort the list by monetary value, break the list into quantiles But, what is the right measure? total lifetime spending average dollars per order The quantile becomes the M component of the RFM cell assignment Now, imagine a customer who has not placed an order in a little while doesnt purchase that often when she does place an order, it tends to be big

275 From Sorted List to RFM Buckets 341 1 2 3 1 4 5 Recency 2 1 3 2 4 3 5 4 5 Frequency

Monetary 276 RFM Cube Recency = 5 511 521 531 541 551 Recency = 4 411 421 431 441 451 Recency = 3 311

321 331 341 351 Recency = 2 211 221 231 241 251 Recency = 1 111 121 131 141 151 Frequency = 5

Frequency = 4 Frequency = 3 Frequency = 2 Frequency = 1 $= $= $= $= $= 5 4 3 2 1 277 RFM Cells are Not All the Same Size

211 221 212 222 111 121 112 122 Consider the small example on left Do the numbers make sense? 15.0% 10.0% 10.0% 15.0% 5.0% 20.0%

20.0% 5.0% 278 Each Cell Has a Different Response Rate RFM CELL PROPORTION RESPONSE 111 112 121 122 211 212 221 222 10.0% 20.0% 15.0% 5.0% 15.0% 5.0% 10.0% 20.0% 3.09% 2.38%

2.71% 2.00% 1.70% 0.99% 1.33% 0.62% The better quantiles along each axis generally have higher response As expected, the 111 cell has the best response and cell 222 has the worst 279 Using RFM Cells to Determine Which Customers Should Receive a Catalog Given a customer list of one million Build a cube of 5 x 5 x 5 = 125 RFM cells Test mail to a random sample with all cells represented Given a budget that will cover test mail plus sending out 100,000 catalogs Mail catalogs to customers in top cells with respect to expected response to reach 100,000 280 RFM is a Testing Methodology RFM involves testing the marketing environment

in order to find the best cells direct test historical data Such testing makes good marketing sense However, it does not use a test set concept, the way we do in building a predictive model 281 Choosing Profitable RFM Cells Calculate break-even response rate Example cost per piece = $1.00 revenue per response = $40.00 net revenue per response = $39.00 let r = response rate profit = 39r (1 r) profit > 0 39r > 1 r 40r > 1 r > 2.5% mail to any cell with > 2.5% response in test mailing 282 Choosing Profitable Cells RFM CELL PROPORTION RESPONSE 111 112 121 122 211

212 221 222 10.0% 20.0% 15.0% 5.0% 15.0% 5.0% 10.0% 20.0% 3.09% 2.38% 2.71% 2.00% 1.70% 0.99% 1.33% 0.62% If we need a 2.5% response rate to break even, we choose cells 111 and 121 These account for 25% of the population 283

Using RFM Cells as Control Groups VCS wanted to test a new catalog designed to appeal to 53-71 year olds The motivation was to exploit nostalgia for teenage and young adult years The new catalog included color images of products black and white images of 1950s and 1960s The test mailing compares target segment against others from the same RFM cell 284 Using Neural Networks for Response Modeling RFM is our baseline method RFM has a number of drawbacks At this time, we seek to gain a basic understanding of neural network models What are the pitfalls to be avoided? In this section, we discuss the neural network response model built by VCS 285 Problems with the RFM Approach RFM cell assignments are not scores is cell 155 better than cell 311? a test mailing is required RFM approach misses important segments Christmas shoppers: in November they may look like they have stopped buying from catalog there may be pockets of good responders in bad

RFM cells there may be pockets of bad responders in good RFM cells 286 Problems with the RFM Approach -continued Proliferation of cells there can be a large number of RFM cells RFM approach hits the same customers over and over the same people end up in the best RFM cells these customers suffer from offer fatigue RFM approach does not make use of some valuable data VCS knows where customers live (demographics) VCS knows about past behavior (purchases, returns, categories) VCS knows which customers are gift givers 287 Problems with RFM Approach -- continued Focus on cells rather than individuals cell-based marketing treats everyone who falls into the same basket the same way because cell definitions tend to be simplistic, this doesnt always make sense Predictive models are more powerful than profiles cell definitions are profiles profiling assigns a customer to a cell whose aggregate behavior is measured

predictive models use input variables to predict future behavior of each customer 288 Back to VCS After some initial success with RFM in the 1980s, VCS became less enchanted with the technique In 1998, they were ready to try something new VCS worked with SAS consultants to see if Enterprise Miner could achieve better results They compared RFM, regression, neural networks, and decision trees 289 Experimental Design The model set data came from previous catalog mailings with data on responders and nonresponders Using this data, VCS built RFM, regression, neural network, and decision tree models to predict who would respond to two other mailings that were also in the past, but more recent than the model set data October mailing November mailing Calculate expected response, revenues, and profit assuming models had been used 290 Available Data on Each Customer Household ID Number of quarters with at least one order placed

Flags indicating payment by personal check, credit card, or both Number of catalogs purchased Number of days since last order Number of purchases from the various catalog types Dollars spent per quarter going back several years Note that R, F, & M are each represented 291 How Models were Judged Compare regression, neural networks, and decision trees with RFM baseline with respect to percent increase in dollar sales Look at return on investment in data mining project How do we guess sales since the comparison is hypothetical? SAS and VCS were very clever 292 Using Historic Response Data to Compare the Models Challenger models will pick a different mailing list than the one generated using the RFM approach How do we guess what someone would have spent had we sent her a catalog, given that we didnt? Each model was used to score the list and choose its own top 1.4 million prospects for actual recipients, use actual dollars spent for non-recipients, use average spending of recipients with the same score

293 The Results The neural network model was the model selected by VCS The model predicted an increase in sales ranging from 2.86% to 12.83% perhaps, depending on the specific catalog The model yielded a return on investment of 1,182% More detailed information is confidential Lets revisit the topic of neural networks 294 Neural Network Architecture output layer hidden layer input layer inputs are usually columns from a database 295 Inside Each Hidden Node output Inputs are real numbers between -1 and 1 (data is transformed) Links have weights

w1 i1 w2 i2 inputs At each node, weighted inputs are summed and passed through a transfer function w3 i3 The output value is between 0 and 1 The output node functions similarly 296 Training a Neural Network Given a pre-specified training set and random initial link weights observe one pattern (a single inputs/output combination) at a time iteratively adjust the link weights stop when final weights best represent the general

relationship between inputs/output combinations This is called backpropagation An alternative approach is to use genetic algorithms 297 Procedure for Using Neural Networks Transform all the input variables to be between -1 and +1 and output variables between 0 and 1 Divide the data into training, testing, and validation sets Train the neural network using the training set until the error is minimized or the link weights no longer change Use the test set to choose the best set of weights for the model Compare models and predict performance using 298 the validation set Neural Networks: Pitfalls to be Avoided Decision trees can easily handle hundreds of input variables Neural networks cannot more input nodes more link weights to be adjusted the larger the training set computationally burdensome one solution: build a decision tree first and use the variables that appear high in the decision tree in the neural network Categorical variables

e.g., re-renter of Ryder trucks in the training set, define a variable that takes on the value 1 if the customer is a re-renter and 0 otherwise 299 Neural Networks: Pitfalls to be Avoided-Continued Be on the lookout for data skew caused by large outliers e.g., in looking at net worth, Bill Gates is assigned 1 and everyone else is assigned 0 differences between cardiologists, college professors, and the homeless would be obscured possible solutions: throw out outliers or perform transformations using logarithms or square roots 300 Final Thoughts on CRM Cross-sell models: see Chapter 10 in Berry and Linoff (Wiley, 2000) Data visualization: application to college selection CRM recap 301 CRM Sources The vast majority of these CRM slides has been borrowed, adapted, or reworked from one of the two sources below: 1. Michael Berry and Gordon Linoff, Customer Relationship Management Through Data Mining, SAS Institute, 2000

2. Michael Berry and Gordon Linoff, Mastering Data Mining, John Wiley & Sons, 2000 I have also consulted Data Mining Techniques (Wiley, 1997) by Berry and Linoff, in preparing these slides 302

Recently Viewed Presentations

  • Neil Armstrong - brooker.mysdhc.org

    Neil Armstrong - brooker.mysdhc.org

    BEING A BORN BABY Neil Armstrong was born on August 5, 1930 to Viola and Stephen Armstrong. WAR TIME In 1951, Neil went to the Korean War. He has accomplished everything. MARRIAGE Neil married Janet Shearon in 1956 BEING A...
  • Chemistry 12.2

    Chemistry 12.2

    Then use the molar mass to calculate the number of moles in that mass. In a typical stoichiometric problem: 1) Convertgrams to moles ... 1 mol Mg. x. molar mass. Grams to Grams (Calculation #4) How many grams of H....
  • Dimensional Analysis

    Dimensional Analysis

    Ratio of Pressure forces to Inertial Force An important parameter in Aerodynamics In cavitation studies, Dp(see formula for Cp) is taken as p - pv where p is the liquid pressure and pv is the liquid vapour pressure, The Cavitation...
  • Case Study 1: UNIX and LINUX - University of Crete

    Case Study 1: UNIX and LINUX - University of Crete

    Case Study 1: UNIX and LINUX 10.1 History of unix 10.2 Overview of unix 10.3 Processes in unix ... The ls Command Steps in executing the command ls type to the shell Flags for Linux clone Bits in the sharing_flags...
  • Applying Geostatistical Methods to Lattice Data: An Initial

    Applying Geostatistical Methods to Lattice Data: An Initial

    Applying Geostatistical Methods to Lattice Data: An Initial Examination of U.S. Presidential Elections in Iowa A.C. Thomas Statistics 225 December 14, 2004
  • ASAS Spacing - Concept and Results

    ASAS Spacing - Concept and Results

    Primary arrival route from North West (shown below) Secondary routes (not shown) merge at ELB and then with this route at GRO ASAS Spacing - Concept and Results Results: ASAS Spacing - Concept and Results ASAS Spacing - Concept and...
  • Measuring Discipline Disparities Mark McKechnie, MSW Senior Consultant

    Measuring Discipline Disparities Mark McKechnie, MSW Senior Consultant

    Equity in School Discipline. This program is a collaboration between the Center for the Improvement of Student Learning and the Office of Equity and Civil Rights to provide technical assistance directly to Washington districts in order to reduce the disparate...
  • 1.1 - Introduction to Cells - St Leonard&#x27;s College

    1.1 - Introduction to Cells - St Leonard's College

    primary (1°) consumer, first-order consumer or herbivore. 3rd TL = secondary (2°) consumer, second-order consumer or carnivore. FOOD CHAINS & WEBS. Food chains represent a linear relationship between organisms. Food webs represent . interconnected .