Customer Retention: how to measure it, build it and keep it.

San Francisco DMA March 16, 2006 3:00 5:00

Prof. Rushen Chahal

How a Modern Database Works

Customer Transactions

Marketing Campaigns

Inputs from retail, phone, web

Marketing Database

Analytic & Campaign Software

Marketing Staff Access By Web

Data Cleaning Standardization Website Appended Data

Modeling & Analytics

Why retention is important: long term loyal customers

Buy more per year Buy higher priced options Buy more often Are less price sensitive Are less costly to serve Are more loyal Have a higher lifetime value

How to retain them

Recruit the right customers to begin with Once you have them, segment them by lifetime value Communicate with them to build loyalty

What proves that communications work?

Manufacturer of indoor lighting products Catalog sent to 45,000 contractors Previous policy: wait for the orders Test: pick 1,200 customers, split into test of 600 and control of 600 Two person pilot program build relationship with test customers to see the results

Change in the number of orders

112%
120% 100% Change in number of orders 80% 60% 40% 20% 0%
1 2

82%

Control vs Test Groups

Change in the Average Order Size

114%
120% 100% Change in average 60% order size 40% 20% 0%
1 2

86%

80%

Control vs Test Group

Total revenue gain: $2.6 million dollars

127% 140% 120% 100% Change in total revenue 80% 60% 40% 20% 0%
1 2

70%

Control vs Test Group

Communications work!
Building

a relationship with customers can be highly profitable Using a database to recreate the old family grocer is a winning strategy Relationship marketing is the way to go

But, with millions of customers

Which ones should you spend resources on? If you communicate with everyone, you will not have enough resources to retain the very best. To select the best, you need to compute customer lifetime value

Lifetime Value

Why we need Lifetime Value Analysis

We need to know the value of our customers, so as to properly target our sales and retention efforts We need to discriminate among our customers to acquire and retain the best

Lifetime Value Analysis Goal: Determine...

where to put your retention dollars the value of each retention strategy where to put your acquisition dollars how much to spend on acquisition

What is lifetime value?

Net present value of the profit to be realized on the average new customer during a given number of years. To compute it, you must be able to track customers from year to year. Main use: To evaluate strategy.

Examples of Profitable Strategies

User

Groups Newsletters Surveys and Responses Loyalty Programs Customer and Technical Services Membership cards and status levels Event Driven Communications

Event driven communication:

Dear Mr. Hughes:
Ridgeway Fashions Leesburg, VA 22069

I would like to remind you that your wife Helenas birthday is coming up in two weeks on November 5th. We have the perfect gift for her in stock. As you know, she loves Liz Claiborne clothing. We have an absolutely beautiful new suit in blue, her favorite color, in a fourteen, her size, priced at $232.00.

If you like, I can gift wrap the suit at no extra charge and deliver it to you next week, so that you will have it in plenty of time for her birthday. Or, I can put it aside so you can come in to pick it up. Please call me at (703) 754-4470 to let me know which youd prefer. Sincerely yours,

Robin Baumgartner
Robin Baumgartner, Store Manager

Lets look at a retail operation

Before and after a loyalty program

LTV Before New Strategies

Retention Rate Customers Visits Per Year Spending Per Visit Revenue Year 1 40% 200,000 1.4 $50 $14,000,000 Year 2 45% 80,000 1.6 $60 $7,680,000 49% $3,763,200 $3,763,200 $3,916,800 1.12 $3,497,143 $4,097,143 $20.49 Year 3 50% 36,000 1.8 $70 $4,536,000 48% $2,177,280 $2,177,280 $2,358,720 1.32 $1,786,909 $5,884,052 $29.42

50% Cost Percentage $7,000,000 Costs $6,400,000 Acquisition Cost $32 $13,400,000 Total Costs Profit Discount Rate NPV Profit Cum NPV Profie Lifetime Value $600,000 1 $600,000 $600,000 $3.00

Discount Rate Basic Formula

Market Rate of Interest...5% Assume Risk (Double rate)...10% Years = n Interest = i Formula: D = (1 + i)n Calculation of rate after 2 years: D = (1 + .10)2 = (1.10)2 = 1.21

New Retention Strategies

Provide all customers with a card or register their credit cards Birthday Club Communicate with them Give them premiums if they shop a lot Lets see what could happen

With New Strategies

Retention Rate Customers Visits Per Year Spending Per Visit Revenue Cost Percentage Costs Acquisition Cost $32 Database Costs Loyalty Program Loyalty Costs Total Costs Profit Discount Rate NPV Profit Cum NPV Profie Lifetime Value Year 1 50% 200,000 1.6 $55 $17,600,000 50% $8,800,000 $6,400,000 $500,000 $5.00 $1,600,000 $17,300,000 $300,000 1 $300,000 $300,000 $1.50 Year 2 60% 100,000 2 $70 $14,000,000 49% $6,860,000 $250,000 $8.00 $1,600,000 $8,710,000 $5,290,000 1.12 $4,723,214 $5,023,214 $25.12 Year 3 65% 60,000 2.4 $80 $11,520,000 48% $5,529,600 $150,000 $10.00 $1,440,000 $7,119,600 $4,400,400 1.32 $3,333,636 $8,356,851 $41.78

Effect of adoption of new strategies

Year 1 $3.00 $1.50 -$1.50 -$300,000 Year 2 $20.49 $25.12 $4.63 $926,071 Year 3 $29.42 $41.78 $12.36 $2,472,799

Old LTV New LTV Change With 200,000 members

What is the proper computation period?

Which is the correct lifetime value? 1, 2, 3, 4, 5 or more years? They are all correct. Which you use depends on your product or service. Long lifetimes: banks, insurance, utilities. Short lifetimes: discount houses, package goods, catalogers.

Five Ways to Boost LTV with Database Strategies

Increase the retention rate Increase the referral rate

Increase the spending rate Decrease the direct costs Decrease the marketing costs

Compute a base lifetime value Dream up a new strategy. Estimate the benefits and costs Determine whether your new lifetime value goes up or goes down Dont undertake any new strategy until you can prove it will be successful

How to use lifetime value

Using lifetime value to get budget approval

Database marketing budgets are usually carved from somewhere else You have to prove that you will make better use of the funds than the others Lifetime value can supply testable numbers that CFOs can understand Base your budget on solid numbers backed up by valid tests

What your new budget will buy

Old LTV New LTV Change With 200,000 members

Year 1 $3.00 $1.50 -$1.50 -$300,000

Year 2 $20.49 $25.12 $4.63 $926,071

Year 3 $29.42 $41.78 $12.36 $2,472,799

How you got there

Retention Rate Customers Visits Per Year Spending Per Visit Revenue Cost Percentage Costs Acquisition Cost $32 Database Costs Loyalty Program Loyalty Costs Total Costs Profit Discount Rate NPV Profit Cum NPV Profie Lifetime Value Year 1 50% 200,000 1.6 $55 $17,600,000 50% $8,800,000 $6,400,000 $500,000 $5.00 $1,600,000 $17,300,000 $300,000 1 $300,000 $300,000 $1.50 Year 2 60% 100,000 2 $70 $14,000,000 49% $6,860,000 $250,000 $8.00 $1,600,000 $8,710,000 $5,290,000 1.12 $4,723,214 $5,023,214 $25.12 Year 3 65% 60,000 2.4 $80 $11,520,000 48% $5,529,600 $150,000 $10.00 $1,440,000 $7,119,600 $4,400,400 1.32 $3,333,636 $8,356,851 $41.78

Using lifetime value to get budget approval

Database marketing budgets are usually carved from somewhere else You have to prove that you will make better use of the funds than the others Lifetime value can supply testable numbers that CFOs can understand Base your budget on solid numbers backed up by valid tests

Who is going to defect?

Besides LTV, you can develop a model that predicts which customers are most likely to leave. Putting that model with LTV you can refocus your entire retention strategy You create a Risk Revenue Matrix

Focus on A and B: 44% of your customers.

Probability of Leaving Soon High Medium Low Priority A Priority B Priority C Priority B Priority B Priority C Priority C Priority C Priority C

LTV High Medium Low

Who uses LTV in marketing?*

DMA survey shows 52% of Consumer Only marketers use LTV. 25% of B to B use LTV. 49% Larger companies ($100 million or more) use LTV. 32% smaller companies use LTV 65% plan to use LTV more extensively in 2006 70% use LTV to decide when to reactivate a lapsed customer. 68% determine promotions by LTV

*DMA Survey 2005

Conclusion: you can do this

Create a lifetime value table for your customers. Put LTV into each customer record Use LTV to determine your marketing strategy Use it to improve retention, cross sales, and profits

Break

Why you need customer segments

Customers are usually very different College students, senior citizens, families with children, empty nesters The same message to all may not work so well. Solution: create segments, and design a program for each segment.

How one retail store created 9 customer segments.

Segments differ from status levels

Segment Strategy

An ideal segment

Has definable characteristics in terms of behavior and demographics: for example, Retired Couples Is large enough in terms of potential sales to justify a custom marketing strategy with appropriate rewards and budget Has members who can be motivated by cost effective rewards to modify their behavior in ways that are profitable for your company Makes efficient use of available data to support segment definition and marketing efforts Can be measured in performance, with control groups Justifies an organization devoted to it: can be a single person, or part of a persons time, but there should be someone who owns each segment.

A valid segment strategy involves:

Communications to the segment (direct mail, email, on-location personal attention) Rewards designed to modify behavior Controls to measure the success of the strategy A budget for implementation of the strategy Specific goals and metrics for engagement: for behavior modification An organization that accepts responsibility for the segment

Segment action plan:

A roadmap showing what will happen when. Send each policyholder a birthday card and a policy review 45 days before their policy renewal date. A budget for the infrastructure and for the segment marketing plans An organization chart that shows who is responsible for each segment Specific goals to be achieved with milestones for measurement of success

Using Clusters as segments

How one non profit measured success by cluster- Their best

Their worst in terms of response and contributions

Success from mailing only to the best, and not mailing the worst

$5 Million more in net gross revenue.

Multi-channel users are more loyal

Retention Rate Year 2
Catalog, retail, web Catalog & Web Retail and Web Retail & Catalog Web Only Catalog Only Retail Only
0% 10% 20% 30% 40% 50% 60% 70% 80%

Illustrative numbers from several case studies

Why the web is important to retention

Web customers are more affluent Their average order size is 12% higher than phone orders. The cost of the web order is 16% lower than phone orders. Typical incentive offered is 5% off on any order over $50. Result: 11% of non web customers shift to the web every year.

Creating a club on the internet

A company selling sporting goods created an internet member club. When DB was built they learned that:

Club members bought 11 times more than non club members. In two years, 81% of club members became multi-buyers. The club boosted retention

Club Members Retention

80.5%
90.0% 80.0% 70.0% 60.0% 50.0% 40.0% 30.0% 20.0% 10.0% 0.0%

Conversion to MultiBuyers after two years

23.4%

Goal Club

Non Goal Club

Cataloger Customer Retention

Miles Kimball sent 20,000 emails with three different catalogs, and 20,000 with the three catalogs alone. Those who got the emails bought 18% more than those who got the catalogs alone.
118
120

115

110

105

100

100

95

90

Control

Test

More sales = Higher overall retention levels

Retailer Customer Retention Video retailer sent email newsletters to 170,000

customers for 6 months. Control group of 14,000 got no emails Retail sales to test group was 28% more than to those without emails.
128
140

100
120 100

80

60

40

20

Control

Test

More sales = Higher overall retention levels

One Click Ordering

With the web we use cookies to say, Welcome back Susan. We keep her credit card on file if she wants so she can do one click ordering Result, compared to controls, is higher retention and annual revenue from those who have one click ordering available.

Tests and controls

Essential to measuring the effectiveness of retention programs

Why controls are essential

The sales force acquires new customers Database marketers create higher retention rates How do you prove this? Retention program effectiveness can only be measured using control groups

Every marketing promotion should always be a test

Test those who get the promotion against the performance of those who do not get the promotion If you are sending birthday cards or a newsletter, select 50,000 who do not get birthday cards or the newsletter. Look at the controls spending rate, and retention rate. If there is no difference, your cards or newsletters are a waste of money.

What to measure

Attrition and retention of both groups Migration upward and downward Incremental sales per program and per season Frequency of purchases Dollars spent per trip and per season Number of departments shopped Number of items purchased Share of customers wallet

Illustration: Birthday Gift

Get customers to record their birthdays with their emails. On their birthday, send them a birthday Pizza Coupon One fast food restaurant offered a $10 birthday coupon to 215,000 customers. Of the coupons sent out, 86,612 were redeemed ($866,120) producing overall sales of $2,900,000 a sales increase of $2 million.

Live Agent
74% of shopping carts abandoned at checkout. Reason: customers have some question. They are unsure about the product, service, color, delivery, etc. Solution: put a live chat button at checkout time. Have live agents available to answer questions. Result: increased retention and sales

Caller ID
Use Caller ID to bring customers complete purchasing history on the screen before the agent begins talking. Result, she can talk to the customer as if she knew her. Result: Increased retention. Greater opportunity for cross sales.

What should you do to keep your customers?

Select loyal customers to begin with. Reward agents for customer loyalty. Set up a customer communications plan Calculate LTV of each customer Use modeling to predict churn and to determine the Next Best Product Combine LTV and NBP to run a proactive retention program

Optimize your inducements