The customer loyalty solution

Required reading!” —Eric Webster, Assistant Vice President, Customer Marketing, State Farm Insurance “Provocative, stimulating, interesting, and loaded with case studiesfrom dozens of co

The Customer Loyalty Solution

“Arthur blows past the CRM hype and lays out the best of databasemarketing, presenting case study after case study of how to do itright (and sometimes not so right!) His integration of current marketing strategies, database marketing techniques, and how theInternet really helps database marketers provides new insights thateveryone will learn from Required reading!”

—Eric Webster, Assistant Vice President, Customer Marketing, State Farm Insurance

“Provocative, stimulating, interesting, and loaded with case studiesfrom dozens of companies, Arthur’s book should be the bible foranyone doing cutting-edge database marketing today.”

—Mike Brostoff, Chairman, CSC Advanced Database Solutions

“An interesting and entertaining read which provides practical insights into the evolving world of Database Marketing/CRM.Arthur's ability to bring focus to the day-to-day challenges (and misconceptions) of the discipline makes the book a great resourcefor the experienced database marketer as well as the novice

—Kay M Madati, Relationship Marketing Manager,

BMW of North America, LLC

The American Economy, 1968

The American Economy, 2d ed., 1969

The Complete Database Marketer, 1991

Strategic Database Marketing, 1994

The Complete Database Marketer, 2d ed., 1996

Don’t Blame Little Arthur; Blame the Damn Fool Who Entrusted Him with the Eggs, 1999

Strategic Database Marketing, 2d ed., 2000

Customer

Loyalty

Solution

What Works (and What Doesn’t)

in Customer Loyalty Programs

Arthur Middleton Hughes

McGraw-Hill

New York Chicago San Francisco Lisbon London

Madrid Mexico City Milan New Delhi San Juan

1976, no part of this publication may be reproduced or distributed in any form or by anymeans, or stored in a database or retrieval system, without the prior written permission of thepublisher

0-07-142904-2

The material in this eBook also appears in the print version of this title: 0-07-136366-1

All trademarks are trademarks of their respective owners Rather than put a trademarksymbol after every occurrence of a trademarked name, we use names in an editorial fash-ion only, and to the benefit of the trademark owner, with no intention of infringement ofthe trademark Where such designations appear in this book, they have been printed withinitial caps

McGraw-Hill eBooks are available at special quantity discounts to use as premiums andsales promotions, or for use in corporate training programs For more information, pleasecontact George Hoare, Special Sales, at george_hoare@mcgraw-hill.com or (212) 904-

4069

TERMS OF USE

This is a copyrighted work and The McGraw-Hill Companies, Inc (“McGraw-Hill”) andits licensors reserve all rights in and to the work Use of this work is subject to these terms.Except as permitted under the Copyright Act of 1976 and the right to store and retrieve onecopy of the work, you may not decompile, disassemble, reverse engineer, reproduce, mod-ify, create derivative works based upon, transmit, distribute, disseminate, sell, publish orsublicense the work or any part of it without McGraw-Hill’s prior consent You may usethe work for your own noncommercial and personal use; any other use of the work is strict-

ly prohibited Your right to use the work may be terminated if you fail to comply with theseterms

THE WORK IS PROVIDED “AS IS” McGRAW-HILL AND ITS LICENSORS MAKE

NO GUARANTEES OR WARRANTIES AS TO THE ACCURACY, ADEQUACY ORCOMPLETENESS OF OR RESULTS TO BE OBTAINED FROM USING THE WORK,INCLUDING ANY INFORMATION THAT CAN BE ACCESSED THROUGH THEWORK VIA HYPERLINK OR OTHERWISE, AND EXPRESSLY DISCLAIM ANYWARRANTY, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TOIMPLIED WARRANTIES OF MERCHANTABILITY OR FITNESS FOR A PARTICU-LAR PURPOSE McGraw-Hill and its licensors do not warrant or guarantee that the func-tions contained in the work will meet your requirements or that its operation will be unin-terrupted or error free Neither McGraw-Hill nor its licensors shall be liable to you or any-one else for any inaccuracy, error or omission, regardless of cause, in the work or for anydamages resulting therefrom McGraw-Hill has no responsibility for the content of anyinformation accessed through the work Under no circumstances shall McGraw-Hilland/or its licensors be liable for any indirect, incidental, special, punitive, consequential orsimilar damages that result from the use of or inability to use the work, even if any of themhas been advised of the possibility of such damages This limitation of liability shall apply

to any claim or cause whatsoever whether such claim or cause arises in contract, tort orotherwise

DOI: 10.1036/0071429042

and her two brothers and sisters-in-law,

Fernando Errázuriz Guzmán and María Eugenia Oyarzún José Miguel Errázuriz Guzmán and Mónica Lopez, who gave me the opportunity to write this book in Chile in 2002

How to Touch a Customer’s Life 3

Combining the Database with the Web 6

Computing Lifetime Value 9

Use of Database Marketing for Acquisition 12

Assumption 1: Why People Buy Things 23Assumption 2: Timely and Relevant Offers 24Assumption 3: Becoming Customer-Centric 25Assumption 4: CRM Mathematics 26

vii

Copyright 2003 by Arthur Middleton Hughes Click Here for Term of Use.

The Loyalty Effect 29What to Do with the Data 32

The Web Is a Passive Medium 41The Web Is an Ordering Tool, Not a Selling Tool 42Web Advertising Was Highly Overrated 43Why Web Supermarkets Failed 43

So, What Is Left? A Huge Research and Transaction Tool 44

Pure-Play Selling on the Web 48Brooks: Clever Efforts to Promote Web Sales 48

Business-to-Business Web Transactions 52Selling on the Web: What We Have Learned 53Measuring the Value of a Site 57

Using LTV to Compute a Name Value 90

Calculating LTV by Segments 93Increasing the Retention Rate 93The Value of an Email Name 95

Business-to-Business Relationship Building 109Consumer Relationship Building 110Keeping the Advertisers 111What Has Happened to RFM? 115

Do Communications Change Behavior? 116

Promoting Music Using Emails 119Emails from Racing Fans 128

Constructing an Email Test 132

C HAPTER 8 MARKETING TO CUSTOMER

Reverse Phone Matching 167

Targeting Customer Segments 173

Reaching Mothers-to-Be 178How Many Segments Should You Create? 180Summary of Case Study Results 183

Supply Chain Management 204Panduit’s Vendor-Managed Inventory 208FedEx Shipping Partnerships 209

Loyalty Programs 220Automatic Trigger Marketing 227Building Loyalty for a Lottery Program 231

Profitable Use of Customer Data 238Dealing with Churn in Cellular Service 239Getting Customers to Join a Club 241

Super Customer Service 250

Customer Service on the Web 254

Letting the Supplier Come behind the Counter 258

Use of Database Marketing 271

Use of Collaborative Filtering 280What Catalogers Have to Guard against When They

Lifetime Value Changes for Catalogers 282Building B2B Catalog Sales 285

C HAPTER 13 FINDING LOYAL CUSTOMERS 293

Finding Leads for Equipment Financing 296Managing Leads on the Web 300Building a Relationship with Prospects 302Telling Them What They Want to Hear 305

Getting Customers to Enter Their Own Profiles 328

Database Marketing Works 338

like religious converts They believed in it.

Inside industry, some companies, including Kraft Foods, AmericanAirlines (and many other airlines), American Express, Hallmark, JCPenney, Neiman Marcus, MCI, and scores of others, formed huge data-bases The idea was that the company could reach its customers directly(instead of through mass marketing) and build their loyalty by directcommunications In most cases, it worked Marketers soon learned how

to set up control groups to prove whether what they were doing wasworking They became skillful at what they were doing

At first, these large databases were maintained on mainframes, withsmaller databases kept on midrange computers With the advent of the

PC and servers, however, all large databases moved to servers, and smalldatabases were maintained on PCs Companies that kept their data onmainframes found that they were having considerable trouble keeping

up with the twists and turns of the database marketing revolution

One twist was the need to create a relational database, which is theoptimum format for database marketing What that means is that for

Copyright 2003 by Arthur Middleton Hughes Click Here for Term of Use.

each customer, we want to be able to keep all that customer’s tions and demographics available, organized in tables for easy access anduse in marketing We append external data and add surveys, profiles,and preferences The mainframe programmers just could not keep up.Today, all large databases are maintained on servers using Oracle orSQL Server There are also a number of good database software pack-ages for PCs.

transac-The advent of the Web has made the greatest change in databasemarketing since it began The Web has meant that

• Marketers can access databases directly from their desktop PCs,using the Web

• Many companies have Web sites where customers can view andorder products

• Customers can contact companies not only by calling toll-freenumbers, but also through the Web, which saves companiesmillions of dollars

• Companies can communicate with their customers frequently atalmost no expense by using email The result is many morecommunications, leading to increased loyalty and sales

• Business-to-business use of the Web has exploded, leading in somecases to vendor-managed inventory, in which suppliers keep track

of what their customers have in stock and help them to becomesuccessful

There have been a number of wrong turns, which are described inthis book:

• Selling to consumers on the Web proved to be a major

disappointment, although the Web is an excellent ordering andcustomer contact medium

• Customer relationship marketing (CRM) based on the idea of to-one marketing using a massive data warehouse, which manythought would be an advanced method of database marketing,proved to be an expensive failure

one-• Packaged goods manufacturers discovered that there was no profit,and in fact considerable losses, from maintaining a database of

ultimate customers The increase in sales from personal

communication did not pay for the expense of the messages and

the database

• Marketers learned that they should use their databases to create

customer segments and treat each segment differently Treating

all customers alike was a loser, since customers differed markedly

in their profitability

Valuable lessons have been learned The purpose of this book is todescribe what works and what doesn’t work in database marketing Toomany books describe only the success stories without talking about thefailures In each chapter of this book, you will find a list of what worksand what has failed to work

Marketing to Customer Segments Gold customers Lifetime value

Web profiling Penetration analysis Next best product

CRM data warehouses

Treating all customers alike

Figure I-1 What Works and What Doesn’t Work

The book contains more than 40 cases from practitioners out the United States, Canada, Australia, and Norway I chose these par-ticular cases because each of them describes the difficulties that had to

through-be overcome and the way success was measured Each one makes apoint Probably the key point is that success does not come from sim-ply doing the following:

• Building a database

• Appending data

• Buying expensive software

• Running models

• Communicating with customers

Success comes from developing intelligent strategies that build tomer loyalty and repeat sales To be successful, database marketers have

cus-to think like cuscus-tomers and say, “Why would I want cus-to be in that base? What is in it for me?”

data-Then they have to dream up strategies that they think will build tomer loyalty They have to test these strategies on a small scale, usingtest and control groups They have to be constantly coming up withnew ideas because they are competing with other marketers who willcopy anything that works and destroy its novelty

cus-Database marketers constantly work with tiny response rates Ofcourse, everyone has heard of marketers who have gotten responses of

20 or 30 percent, but that is very rare Instead, 2 percent is a goodresponse rate for a direct-mail offer, and 1 percent is a good responserate for an email offer Some marketers get up every day and go to workknowing that they will not get better than 4⁄10of 1 percent response onanything that they do, but they keep on going because in many casesthat rate is enough for profitability

Incrementalism

One of the first things to remember about database marketing is thatits benefits are incremental Every company that uses database market-

ing successfully has another profitable sales channel already working for

it, such as retail stores, mass marketing, catalogs, or sales through ers or agents The company uses the profits made through these otherchannels to build the database and develop the communications thatmake it work The idea in database marketing is that you can increaseyour profits by an incremental amount through building customer rela-tionships and information that have the result of

deal-• Increasing the customer retention rate or repurchase rate

• Increasing referrals

• Increasing the cross sales and up-sales

• Using the customer profile to find more loyal and responsive

prospects

The database does not have to bear the full responsibility for sales—that is done by the other channels What the database does is to increasethose sales by a certain percentage The incremental amount is seldommassive—perhaps 5 to 10 percent is all But that 5 percent can be veryprofitable

Marketers have learned how to measure profitability and lifetimevalue There are many examples of this in this book Marketers havelearned to determine who their Gold customers are and how to retainthem They have learned to use recency, frequency, monetary (RFM)analysis to determine which customers are most likely to respond tooffers They have learned many rules, such as the following:

• Giving customers a choice in a direct offer always reduces

response

• Don’t market to Gold customers, work to retain them

• Don’t treat all customers the same way—save your resources to

reward your best ones

• Always set up control groups to measure your success—if you

don’t, you can’t prove that what you have done has really worked

• Compute the lifetime value of your customers and put it into the

customer records so that you can use it to evaluate the success of

your marketing strategy

• The more products you sell to customers, the higher their

retention rate will be

• Customers like to receive communications Almost any

communication will increase sales and retention rates

• Referred customers have higher lifetime values and retention ratesthan the average customer

• Offering discounts to attract new customers produces disloyalcustomers

• Use the Web and email as a powerful two-way customer contactmedium

You will have fun with this book—reading the chapters, taking thequizzes, and using the charts All the numerical charts in this book can

be downloaded free from www.dbmarketing.com, where you will find theanswers to the quizzes, more than 100 magazine articles, free software,and a lot more

If done properly, database marketing will be fun for both the customerand the marketer It is an intriguing game of communication and life.Enjoy it while you increase your sales and profits

I will be glad to hear from any readers on any subject at any time Youmay reach me at dbmarkets@aol.com Point out mistakes in the book,ask for help with marketing, or send me a case study that I can use in afuture book

Arthur Middleton Hughes

Vice President for Business DevelopmentCSC Advanced Database Solutions

2100 South Ocean Drive, Suite 16AFort Lauderdale, FL 33316dbmarkets@aol.comNovember 2002

ACK NOWLEDGMENTS

xix

In addition to the many people mentioned in this book, I want to ognize the special contribution of several individuals who have beeninstrumental in providing me with the ideas that have shaped my knowl-edge of database marketing:

rec-• Paul Wang, Associate Professor of Database Marketing at

Northwestern University Paul and I gave 28 two-day seminars

together from 1994 to 2000 Paul is an outstanding teacher and a

great friend

• Bob McKim and Evelyn Schlaphoff at msdbm in Los Angeles,

where I worked for 2 years Bob and Evie taught me a great deal

about the use of the Web for database marketing Their ideas

illuminate these pages

• Fredrick Reichheld, Professor at Harvard Business School, whose

book The Loyalty Effect completely revised my thinking in a

number of crucial areas

• Beth Clough, Director of Partner Relations at msdbm, a good

friend and an expert on database marketing, who provided valuablehelp in the editing of this book

• Mike Brostoff, CEO of CSC Advanced Database Solutions in

Schaumburg, Illinois, who has run this highly successful database

company for 20 years Mike and Jeff Lundal of CSC invited me towork there as vice president for business development I have beenthrilled to join their team to spread the good word on database

marketing throughout the world

Copyright 2003 by Arthur Middleton Hughes Click Here for Term of Use.

This is my fifth book on database marketing My first, The Complete

Database Marketer, was published in 1991 As I write each

subse-quent book, I am constantly amazed at the changes that have takenplace in the techniques and the uses of database marketing Today is noexception There is so much that is new and different that we are almosttalking about an entirely new marketing method

How It Began

I like to repeat the story of the genesis of database marketing It got itsstart in a big way in the early 1980s, when mass mailers such as Amer-ican Express and State Farm Insurance started using their customer lists

to build ongoing relationships with their customers after the initial sale,leading to increased retention and cross sales But the roots of databasemarketing go back to a period in the United States before there weresupermarkets

Back in those days, all the groceries in the United States were sold insmall corner grocery stores The proprietors knew their customers’names They would stand at the door and greet their customers by name

as they entered, asking them about their families and their concerns Theyput things aside for customers, helped them carry heavy packages out totheir cars, and built strong and lasting relationships They built their busi-nesses by developing and cultivating the loyalty of their customers

Copyright 2003 by Arthur Middleton Hughes Click Here for Term of Use.

These fellows are all gone today They were forced out of business bythe supermarkets Mass marketing took over, along with mass production.Prices came down Quality, quantity, and variety went up The averagecorner grocer had 800 stockkeeping units (SKUs) in his store The aver-age supermarket today has more than 30,000 SKUs The change affectedthe way Americans lived In 1950 the average American family was spend-ing 31 percent of its household budget on food Today the average family

is spending only 14 percent on food, and the food it is getting is better

in both quantity and quality than the food it was spending 31 percent on

50 years ago Because of the lowered cost of food, families have muchmore money to spend on hundreds of other products that were out of thequestion 50 years ago So we have all gained

At the same time, we as suppliers have all lost We have thousands,hundreds of thousands, or millions of customers, but we don’t know them,and they don’t know us Loyalty has gone out the window If you talk to

an employee in a supermarket, you are interrupting his regular work,which is certainly not standing by the entrance and chatting with the cus-tomers You are loyal to that supermarket until tomorrow’s newspaper,when a certificate from somewhere else leads you to drive to another store

In the mid-1980s, computers came into wide use This developmentenabled merchants to begin to bring back some of the intimacy that pre-vailed in the presupermarket days The software and hardware becameincreasingly sophisticated, and their prices have been in a free fall foryears As a result, it is now possible to keep, economically, in a computerthe kind of information on customers that the old corner grocer used

to keep in his head and to use that to build lasting, profitable ships with customers

relation-Many companies have collected huge amounts of data about ers and their transactions, but they have failed to make profitable use ofthese data There is one principle that has remained true throughout theperiod since 1985: Database marketing is effective in building customerloyalty and repeat sales only if the customer benefits from it The cus-tomer has to think, “I’m glad that I’m in that database because ,” withthe supplier filling in a meaningful end to the sentence If the databasedoes not touch the customer’s life in some way that is satisfying to her,

custom-she will ignore the communications, leave her gold card behind whenshe shops, and refuse to become loyal Too many companies haveignored that principle and, as a result, have failed to succeed in data-base marketing.

How to Touch a Customer’s Life

How do you touch a customer’s life using a database? There are manyways, some of them so simple that we overlook them When I fly intoSan Francisco, check into a Hyatt hotel, and push the button for roomservice, the response is, “Yes, Mr Hughes What can I get you to eat?”

I have been in this hotel for less than 20 minutes, yet down in the kitchenthey already know and use my name! This is possible because Hyatt, likemost other hotels today, has Caller ID on its internal hotel telephonesystem When you push that button in your room, the Caller ID goes tothe database that Hyatt created when you checked in and pulls up theinformation, “Arthur Hughes, Room 1202” on the screen so that roomservice can call you by name This is database marketing It is what theold corner grocers used to do, and it is now made possible by moderntechnology used in a creative way You are 2000 miles from home, yetpeople know you and recognize you It makes you feel great!

Caller ID, of course, is not just for hotels Customer service ments throughout America are using it to recognize customers whenthey make repeat calls With this technique, before a call is answered,the database showing the customer’s complete purchase history, demo-graphics, and preferences is brought up on the customer service rep’sscreen, so that he can respond, “Mrs Webster So nice to hear from youagain How did your granddaughter like that sweater you gave her lastOctober?” This is the kind of thing the old corner grocer used to say,and we can say it today But there is a new twist: Using cookies, we areable to personalize our Web sites so that customers get the same type

depart-of personal greeting when they return to Amazon.com, Barnes &Noble, Staples, Office Depot, and hundreds of other Web sites Data-base marketing today has techniques that make its promise come alive

How We Have Changed

There are scores of different examples in the pages of this book Theprinciples are the same, but the methods have changed considerably inthe past 15 years:

• When database marketing began, major databases were built on amainframe Today they are built on a server using Oracle or SQLServer

• In the beginning, most databases were maintained by the informationtechnology (IT) department or a service bureau and could not beaccessed directly by marketers When access was available, it was

by terminals linked to the database by telephone lines Today,marketers using PCs access databases over the Web

• Data used to enter the databases in batch mode as keypuncherscopied customer responses received through the mail Even tele-marketers entered data onto temporary disks, which were laterused to update the mainframe database in batch mode Todaycustomers place their orders and update their profiles on the Web,through a Web page connected with a database on a server Manytelemarketers also work directly with live data on the same server

• Database reports used to be produced monthly by programmersand sent to the users in hard copy Today, marketers have a menu

of reports on their screens, which they can run every day and printfrom their desktop PC printers

• In the past, to select customers and prospects from a database,marketers would send detailed memoranda to the programmers.Upon receipt of these memoranda, programmers would write com-puter programs to select the desired records and send reports on theresults to the marketers by fax The process used to take a couple

of weeks Today, using E.piphany and other advanced applications,marketers select their own records for promotions using their PCsthrough the Web, without the need for IT assistance The selectedrecords can be automatically downloaded to the marketer throughthe Web or saved for downloading by the mail shop or the e-blast-ing vendor The process takes a few minutes

• Data appending was rare and was done mainly by the largest nies Today, many companies add demographic data (age, income,

compa-presence of children, credit ratings, and scores of other facts) to

their customer database routinely Business-to-business marketers

add Dun & Bradstreet (D&B) data (SIC code, annual sales, number

of employees) to all customer and prospect files These data are usedfor segmentation and crafting of marketing communications

• Customer segmentation was rare Few companies had identified theirGold customers, or knew what to do with them once they were iden-tified Today many companies have segmented their customer base

and have developed different marketing strategies for each segment

• Communication with customers was mostly by direct mail or

telephone calls Phone calls cost several dollars each Letters

were also expensive As a result, communications with customerswere limited to monthly statements and sales calls Today,

because of the Web, scores of additional possibilities for

com-munication with customers have opened up Customers order onthe Web, are thanked for their order, receive a confirming

email, are notified by email when products are shipped, and are

surveyed to determine whether the products arrived and were isfactory All these messages were impossible when database

sat-marketing first began Direct-sat-marketing promotions were sent

out by third-class mail in massive batches They still are But to

this has been added email promotions to customers who have

agreed to receive them These email promotions cost almost

nothing to send, compared to the heav y costs of direct tomer preferences used to be requested through direct-mail or

mail.Cus-telephone surveys These were expensive and, consequently,

rare—they were used mostly for research rather than for

cus-tomized communication Today, customers can, and do,

com-plete their personal profiles via the Web, so that outgoing

communications can reflect their preferences

• Customers used to call toll-free numbers to get information on

product shipments or technical data, and to ask for all sorts of mation that was stored in a company’s internal records They still

infor-do Increasingly, however, companies are making that information

available to customers through the Web This saves the companiesmillions of dollars per year and builds customer loyalty by givingthe customers direct access to company archives.

• Models were rarely understood or used Today all companies withproducts that involve churn, such as credit cards and cellular orlong-distance telephone service, routinely use models to predictand forestall the canceling of service

This is an exhausting list It shows where we were and how far we havecome Perhaps the development that has changed database marketingmost profoundly in the past two decades is the advent of the Web Today

we can do things that we only dreamed about doing 10 years ago We cantruly build customer relationships with almost as many personal com-munications as the old corner grocers used to have with their customers

But Does It Work?

We can do all these things, but do they really produce more profits? This

is an area in which great strides have been made Most companies areacutely aware of the need to use control groups to validate the effective-ness of every single marketing strategy and initiative In this book you willfind more than 40 case studies that were completed in the past couple

of years by companies that have been able to prove that their techniquesare succeeding in building loyalty, increased sales, retention, repurchase,and reactivation Customer communications work Personalized com-munications work even better Now, with the Web, we can afford tosend millions of desired and accepted messages based on data and pro-files in our databases that affect customer behavior in a positive way

Combining the Database with the Web

Perhaps the most interesting shift in database marketing has been thefact that the database and a Web microsite can be located on the sameserver or can be electronically connected so that they seem to be on the

same server What this means is that the database and the microsite arenot two separate entities, but part of a single customer response and com-munication vehicle Cookies are placed on the customer’s computer sothat when she comes to our site, she sees a site that is configured for her,with things that she has requested to see This is particularly important

in business-to-business marketing For example, Dell began the process

by creating Premier Pages for all customers with 400 employees or more.Dell salespeople contact the purchasing officer at each company andmake a volume pricing deal that is hard to refuse: “We will create aPremier Page for your company on the Web If you get your employeeswho need Dell products to use this page with passwords controlled byyou, we will give them special prices that are not available to the gen-eral public In addition, we will send you monthly reports on what youremployees are buying and how much you are saving as a result ofyour volume pricing deal with Dell.”

Thousands of companies are copying the Dell idea, setting up specialWeb pages for their best customers The system builds sales and loyalty

Of course each purchase gets an on-screen thank you and confirmationthat can be printed out on the employee’s PC The employee also gets

a simultaneous confirming email, plus another email when the product

is shipped She gets a final email survey a few days later asking if theproduct has arrived and requesting her opinion of the process This ispowerful database marketing

Web Response

Another impact of the Web on database marketing is in the area ofresponse to communications Before 1996, customers responded tomarketing communications by phone, mail, or fax Now more and more,people are coming to a Web microsite to respond Because of cookies,the Web site knows who they are and greets them by name, just as theold corner grocer used to do

The microsite is not really a Web site, but part of the database Thecustomer enters an order or responds to the survey questions posed inthe communication, and the data go directly into the database The

database thanks the customer for the order, which is passed directly tothe Web ordering system It then sends the customer a confirming email.

A customer who has previously provided credit card information canorder by one click, instead of giving a name, address, and card numberover and over again

Why do customers prefer this system of response? Because theynever get put on hold They can get an instant response in 24/7 timewithout getting a busy signal or the message “Your call is very impor-tant to us Please stay on the line.” Why do companies prefer this type

of response? Because it saves them millions of dollars Each call to a liveoperator costs a company between $3 and $7 per call Each response to

a microsite has a variable cost of less than $0.02

Customer Service

America leads the world in telephone customer service, but we pay ahigh price for it Since the beginning of the database marketing revo-lution, every company provides a toll-free number for customers to callwith any kind of question Operators sitting at PCs or terminals linked

to the customer database and to company data and archives receive thesecalls through an electronic call director As customers ask questions, theoperators key in commands on their keyboards and read the answers offtheir screens to the waiting customers They also note in the customerdatabase that the customer has called, and what the customer called about.All this expensive service is changing as a result of the Web Cus-tomers are being given access through the Web to the same informa-tion that the customer service reps are seeing on their screens Theyare able to enter the same commands and get the same informationwithout ever being put on hold They can search for the status of theirshipment, for the price and description of obscure needed parts andsupplies, for the text of previously published articles and white papers.They can see and print maps, product descriptions, specifications, illus-trations, and prices This service builds customer loyalty and saves thecompany millions of dollars

Computing Lifetime Value

Until 10 years ago, few companies used their databases to compute tomer lifetime value Banks began the process They were able to deter-mine the profitability of every account on a monthly basis, and to roll that

cus-up to compute household profitability The results were used to segmentcustomers in order to learn which ones were useful to the bank andwhich ones were reducing the bank’s profits

I like to use the chart shown in Figure 1-1 It comes from a major bank

in the South that ranked all its customers by profitability It shows that 80percent of the bank’s profits came from the top 5 percent of its cus-tomers The bottom 28 percent lost 22 percent of its profits This is verypowerful information that was just not available before database market-ing came along What was even more powerful was what the bank was able

to do with this information Clearly the top 5 percent of customers had to

be retained The bank’s whole future rested on these people In addition,something had to change in the bank’s marketing strategy for the bottom

28 percent Why try to retain people who are costing the bank money?Not just banks, but insurance companies, business-to-business mar-keters, and scores of other enterprises began to compute numbers likethese and to develop strategies that made profitable use of them Manycompanies went on to determine potential lifetime value by usingcomputer models to determine all the other products that each customer

was likely to buy from the company in the future Each such producthad three attributes:

• The potential profitability to the enterprise if the customer were

to buy the product (based on the customer’s assets and income)

• The probability that the customer would buy the product if itwere offered (based on the purchases of other customers withsimilar lifestyles)

• The cost to the company of selling the customer the additionalproduct

From this analysis came the determination of the “next best product”for each customer (in terms of profit to the enterprise and likelihood ofresponse) This product was identified in every customer’s databaserecord and shown on the screens used by all personnel who had cus-tomer contact This has proved to be a highly successful and significantuse of database marketing

The Retention Rate

The most important number in any lifetime value table is the retentionrate This is the percentage of newly acquired customers in each seg-ment who will still be buying from your company 1 year later Beforedatabase marketing, few companies knew what this number was or what

to do about it I worked with an insurance company whose retentionrate in the first year was only 54 percent During this year, the com-pany made absolutely no effort to make contact with its customers or

to build a relationship with them Lifetime value computation showedthat an increase of 5 percent in this first-year retention rate wouldincrease profits by $14 million (see Table 1-1)

The cost of getting this additional 5 percent increase was estimated

at $36 per customer, or a total of $3 million, for customer tions The beauty of using a database to do lifetime value analysis is thatyou focus on things that you can change by marketing strategy and thatyou can prove the value of each strategy in finite dollar terms

communica-Using database marketing, the company would not have to commit

the $3 million in the hope of getting this increased return It could use

the database to set up test and control groups to see if the

communica-tions would produce the desired results It could select 5000 customers

to receive the communications (at a cost of $180,000) to see if their

retention rate could be increased by 5 percent If the strategy worked,

then the next year the company could apply the successful methods to

all new acquisitions, spend the whole $3 million, and pocket the $14

million (Table 1-1 and all the other tables in this book can be

down-loaded free from www.dbmarketing.com.)

Modeling for Churn

As the economy has grown, more companies have learned to use database

marketing to retain customers Competition in some products,

particu-larly credit cards, cellular phones, long-distance service, and health care,

has become intense The attrition rate, or the rate at which people switch

to another provider, is very high At the same time, statistical modelers

have found that regression analysis can help in predicting which

cus-tomers are most likely to drop the service The models use a combination

of appended data and transaction history to pick out the variables that

pre-cede a customer’s dropping the service Then advanced database

market-ing techniques are used to concentrate on these potential defectors and

Table 1-1 Effect of 5 Percent Increase in the Retention Rate

New customer Second year Second year

get them to change their minds Defection prediction plus lifetime valueanalysis further concentrates the communications on those whom it is mostprofitable to serve The result is risk/revenue analysis (see Table 1-2).

In this analysis, which is discussed later in this book, a concentration

on those customers who are priority A or priority B reduces the task oflowering churn by 56 percent, saving resources and boosting profits.Bruce McDoniel of Summit Marketing used this system with a largeregional bank to find the right price for priority A customer renewals.The goal was to get customers to renew, but also to maximize revenue.Bruce tried three price offers that netted the bank from $8 to $10 Table1-3 shows the result of the test

This simple test enabled the bank to fix the offer price at $9

Use of Database Marketing for Acquisition

Database marketing is also highly useful for acquiring new customers Thefirst step is to apply lifetime value analysis to existing customers to deter-mine the characteristics of the most profitable and the least profitablecustomers This analysis can be used to steer the sales force towardrecruiting the right sort of people in the first place Frederick Reichheld,

in his path-breaking book The Loyalty Effect, showed us that some people

are inherently loyal, whereas others tend to be disloyal If companies

Table 1-2Risk/Revenue Matrix

Probability of leaving soon LTV High Medium Low

Table 1-3 Price Offers

Number Number Price

tested renewing offer Revenue

concentrate on recruiting loyal customers to begin with, they will have

a much easier time retaining their customers once they are acquired

In business-to-business customer acquisition, customers can be scoredusing Dun & Bradstreet data, such as Standard Industrial Classification(SIC) codes, annual sales, number of employees, and other such data Byusing this type of appended data, enterprises can learn which types ofcompanies are most likely to buy which of their products Their acquisi-tions can be accurately targeted

In consumer acquisition, cluster coding has proved useful in many cases.Cluster coding systems, marketed by Claritas and other providers, seg-ment the U.S population into 62 different clusters based on lifestyle, withcatchy names like Shotguns & Pickups, Money & Brains, Furs & StationWagons, Pools & Patios, and Hard Scrabble If you apply these clusters toyour existing customer base, you may discover that your services appealprimarily to 30 percent or so of the 62 possible clusters Marketing to theremaining 70 percent may be a losing proposition By concentrating youracquisition resources on the prospects that are most likely to buy, you canincrease your success and your profits Cluster coding works for only a fewproducts and services, but where it works, it is well worth the money

Two Kinds of Databases

There are really two different kinds of databases in any company that

is engaged in direct marketing of products and services One is an

oper-ational database, and the other is a marketing database (see Figure 1-2).

An operational database is used to process transactions and get outthe monthly statements

• For a cataloger, the operational database is used to process the

orders, charge the credit cards, arrange shipment, and handle

returns and credits

• For a bank, the operational database processes checks and deposits,maintains balances, and creates the monthly statements

• For a telephone company, the operational database keeps track of thetelephone calls made and arranges the billing for them

A marketing database gets data from the operational database, if there

is one These data consist of a summary of monthly transactions But themarketing database also includes much more It gets additional data from

• Preferences and profiles provided by the customers

• Promotion and response history from marketing campaigns

• External sources, such as Experian, Donnelly, and Claritas

• Lifetime value and RFM analysis, leading to creation of customersegments

• Modeling for churn and next best product

The marketing database passes data back to the operational database

It may advise the operational database

• The segment that each customer has been placed in, which may lead

to operational decisions Gold customers, for example, may getdifferent operational treatment

• Expressed customer preferences, which lead to different

operational treatment; e.g., smoking or nonsmoking rooms areassigned automatically

appended data model scores

RFM & lifetime value

marketing database operational

MarketingcommunicationsSales, shipments, payments

Figure 1-2 Operational and Marketing Databases

The operational database is run by IT It is run on accounting ciples, and it balances to the penny, since there are legal and tax aspects

prin-to its data It is audited by external audiprin-tors It contains only currentdata on customers Old data are archived There are no data on pros-pects until they make a purchase

The marketing database is managed by the marketing department

It is usually outsourced It is not run on accounting principles, and itdoes not need to balance There are few legal or tax considerations Itcontains information on current customers, prospects, and lapsed cus-tomers and communications with them It may retain data stretchingback over a period of several years

A data warehouse combines all the data and functions of both an

operational database and a marketing database (plus, in many cases,other functions such as employee data, payroll, product data, produc-tion data, etc.) in one huge database The data warehouse is so big andcomplicated that it needs to be managed by a committee It is often veryexpensive to build and maintain It is seldom very flexible Marketingpriorities usually get short shrift in a data warehouse, as other prioritiesare often given preference

Why Databases Fail

There have been many database marketing failures during the past twodecades, from which companies have learned what does not work A few

of the reasons are

• Failure to test a strategy on a small sample before a rollout Database

marketing is particularly adapted to setting up tests, which can

prove the validity of any idea before serious money is spent on it

Database marketers know this Company management usually doesnot Pushing for quick profits has led many companies to fail with

ideas that could easily have been tested on a small scale first

• Failure to set aside control groups If you are going to spend a million

dollars on a group of customer communications designed to boost

retention or sales, you always should set aside a group that does not

get the communications in order to prove the value of the effort.Many companies still fail to do this The result, of course, can be dis-aster If the economy is going down and your sales are going down,you may assume that your methods are not working If you set aside acontrol group, however, you might find that the sales of your testgroup went down less than those of the control group, proving thatyour communication program worked, even in adverse circumstances.

• Failure to develop strategies that work Database marketing has often

worked so well that some people have thought that it was the base and the software that were producing the miracles It wasn’t.What produced the miracles were strategies that resonated withthe particular audiences represented in the databases The scarcestcommodities in any database marketing operation are imaginationand leadership: the ability to think up ideas that will work and theability to sell those ideas to management and apply them It is ahighly competitive world today Friends and Family, created byMCI 15 years ago, was a wonderfully creative idea for acquiringnew customers It alone made MCI into a giant long-distance car-rier But once everyone else realized what it was and how it worked,

data-it was copied by others, and data-it no longer had the power data-it once had

To succeed, you have to constantly come up with new ideas Somepeople just don’t understand this They think that if they build adatabase and apply expensive software, profits will roll in This isthe reason for the CRM craze that hit the database marketingworld in the previous decade If you read the CRM manuals, italmost sounds as if customer acquisition and retention is a simplematter of building a data warehouse and applying expensive soft-ware to it In general, CRM has been a gigantic failure We shallreturn to this subject in the next chapter

• Focus on price instead of service The main reason why customers

stop using a product or service is not the product or the price, butthe way they were treated by the company providing the product

or service Management always thinks that price is the problem,hoping that through discounts, it can buy some loyal customers.Discounts do not buy loyalty They reduce loyalty They reducemargin They get customers to think about how much they arepaying instead of about how much they are getting A database

should not be a way to give away coupons It should be a way to

avoid having to give away coupons because the customers are so

happy with the communications and services they are receiving

that they will ignore the deep discounts of the competition

• Treating all customers alike Some customers are more loyal and

profitable than others Your database should identify these loyal tomers, and your marketing strategy should work to retain them

cus-and to encourage others to emulate them You should also do

something to reduce the number of worthless customers on your

rolls If you treat all customers alike, you will never improve your

customer mix A retention budget of $1 million divided by 1

mil-lion customers will do very little to change anyone’s behavior

Divide that $1 million among your 100,000 most valuable ers, and you have $10 per customer per year to reward loyalty andencourage retention You may make some progress Some companymanagements do not yet understand this

custom-Summary

Database marketing is a great success story During the last 20 years,thousands of creative and skillful people have developed and appliednew methods for using databases of customers and prospects to acquirecustomers, build loyalty, and increase sales These methods areexplained in detail in this book But database marketing is more than asuccess story It is a philosophy

Looked at from the customer’s point of view, database marketing is away of making customers happy—of providing them with recognition,service, friendship, and information, for which, in return, they will rewardyou with loyalty, reduction in attrition, and increased sales Generatingprofits by creating genuine customer satisfaction is the goal and hallmark

of satisfactory database marketing If you are doing things right, your

cus-tomers will be glad that you have a database and that you have included

them in it They will appreciate the things that you do for them If you candevelop and carry out strategies that bring this situation about, you are amaster marketer You will keep your customers for life, and you will behappy in your work You will have made the world a better place to live in

What Works

• Getting customers to respond to you on the Web

• Creating premier pages for your best business customers

• Providing customer service information free on the Web

• Determining the lifetime value of your customers and using it toevaluate your marketing strategy

• Determining the next best product for each customer and putting

it in the database for your employees to see and use

• Learning the retention rate for each of your customer segmentsand working to improve it

• Sending email communications to customers who have asked forthem

• Using cookies to recognize and greet customers when they come

to your Web site

• Using risk/revenue analysis to concentrate your marketingresources on customers whose behavior you need to modify

What Doesn’t Work

• Building a huge data warehouse for the purpose of buildingprofitable relationships with your customers

• Treating all customers alike

• Failing to set up control groups

• Failing to test any new idea on a small group first

• Failing to use the Web as part of your database communicationsprogram

c Treating all customers alike

d Email

e Direct mail

2 In the case given in this chapter, it was suggested that the

increased retention was due to

a Customer communications

b Advanced software

c CRM

d More loyal customers

e Cookies on the Web

3 What is considered to be a good average response rate for email

5 In risk/revenue analysis, the marketer concentrates most on

a Those customers in priority A and priority B

b The lifetime value of priority B customers

c The churn model

d The lifetime value model

e The data warehouse

6 What’s wrong with treating all customers alike?

a It’s undemocratic.

b The resources available to devote to the best customers

will be too small

c Even worthless customers deserve a break.

d It is impossible to segment customers by behavior.

e Customers would not like it.