A Primer on the Lifetime Value of a Customer

How do we measure firm performance?

Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

How do we measure firm performance?

How do we measure firm performance? If the firm is composed of several divisions, we also look at profitability of each division.

We most often look at financial performance across time….the income statement. Q1

Q2

Q3

Q4

Total Revenue

9074

10571

32641

7294

Cost of Sales

8004

9269

20220

7288

SG&A

Cement

4226

4933

6642

3491

Total Expenses

12230

14202

26862

10779

Income Before Taxes

-3274

-3877

5603

-3594

Income Taxes

-1309

-1551

2241

-1437

Income after Taxes

-1965

-2326

3362

-2157

Strategic Business Units

Industry Growth Rate

Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

Stars

Question Marks

Cash Cows

Dogs

Operating Margin % Net Operating Assets

Net Operating Assets

1231.9

198.4

379.1

1809.4

228.2

51.4

51.8

331.4

18.5

25.9

13.7

18.3

1098.3

199.9

308.5

1606.7

Profitability by Product May Require Different Line Items

TOTAL

1231.9

198.4

379.1

1809.4

228.2

51.4

51.8

331.4

18.5

25.9

13.7

18.3

1098.3

199.9

308.5

1606.7

Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

Operating Margin %

TOTAL

Low

Lime

Profit before taxes

Profit before taxes

Investments and Other

Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

High

Turnover

Turnover

Lime

Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

1

Profitability/ Square Foot By Category Specialty Foods Staples Wine Produce Meat/Seafood Deli/Prepared 3975 1490 500 930 596 1490 $565.4 $212.6 $135.6 $226.8 $198.5 $981.7

Square Feet Gross Margin

Measuring Firm Performance

Other Total 800 9781 $38.7 $2,359.3

GM/Sq Ft

$0.142

$0.143

$0.271

$0.244

$0.333

$0.659

$0.048

$0.241

Allocated Cost Contribution

404.4 $161.0

233.5 ($20.9)

144.4 ($8.8)

160 $66.8

215 ($16.5)

821.6 $160.1

60.5 ($21.8)

2039.1 $320.2

Contribution/Sq Ft

$0.041

($0.028)

$0.107

($0.014) ($0.018) $0.072

By quarter or year By Strategic Business Unit By Product By Category

($0.027) $0.033

Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

There is one other important way to look at firm performance: Bob

By Customer

• • • •

Phil

Sue

The Big Idea Ravi

Total Revenue

9074

10571

32641

7294

Cost of Sales

8004

9269

20220

7288

SG&A

4226

4933

6642

3491

Total Expenses

12230

14202

26862

10779

Income Before Taxes

-3274

-3877

5603

-3594

Income Taxes

-1309

-1551

2241

-1437

Income after Taxes

-1965

-2326

3362

-2157

Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

Not a New Idea • B2B marketers measure profitability of each large account (customer). Account management generated the need for customspecific P&L statements. • Shop keepers know who their best customers are. • Insurance providers price based on an individual’s (or segment’s) actuarial risk.

The purpose of a firm is to create and satisfy a customer. The customer is King. Love your customers more and your products less If “customer focus” or “customer orientation” is part of your business philosophy, then consider measuring the performance of each customer relationship. Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

But an Idea whose time has come…because • E-business enables direct relationships with customers. – Any firm with a website is forced into relationships with individual customers. – Disintermediation cuts out layers—marketers converse with individual consumers, not just with channel partners.

• Information Technology makes it easier and cheaper to monitor a large number of customer relationships. – Customer clicks are easy to track. – Moore’s law keeps reducing costs.

• Customers expect more – and are less likely to accept “one size fits all.”

Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

2

Love your products less

And your customers more

• How profitable is this product? • How should I price this product? • Are we doing enough new-product development? • How can we extend the product life cycle? • What is our share of market?

• How profitable is this customer? • How should I price this customer? • Are we doing enough new-customer development? • How can we extend the customer life cycle? • What is our share of customer requirements?

Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

Because customers are not alike

Most Valued Customers

Why bother evaluating individual customer relationships?

Second Tier Customers

Below Zeros

Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

Customers to Grow

Customers to Grow

Customers to Reward

Customers to Reward

And should be treated differently

In order to improve firm performance

Customers to Fire?

Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

Customers to Fire?

Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

3

Step 1. Know how many customers you have.

Evaluating Customer Relationships

• Coke and P&G may never know. • Dell knows! Does Compaq? • Some firms that should know do not. – A very large bank knew how many accounts it had, but not how many customers.

• You may have to incent them to identify themselves – frequent shopper cards – online registration – cookies as a last resort Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

Step 2. Develop the capability to construct P&L statements for individual customer relationships.

Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

Customer Interaction Metrics It is important to fully understand the financial implications o f all interactions between your organization and your customers.

• Billing systems need to interoperate with marketing databases. • You may need to build a Customer Database • ABC analysis of the costs of maintaining customer relationships.

Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

for every $400 deposited/month per customer/month per customer/month per transaction per transaction per transaction per transaction

-

promotion newsletter bill renewal notice survey discount/incentive write- offs

+ Added Value: interactions leading to orders, upgrades, conversions, referrals, etc...

- Negative Value: interactions leading to returns, dismissals, credit write offs, bad payment...

The Firm

Source: “The Evolution of Direct Marketing: Enterprise Customer Management,” Presentation by ChristopherBecke(Darden ’92), May 7, 1999.

Deposit revenues

Bank calculated revenues/costs per customer interaction:

$1.00 $6.00 ($3.00) ($2.50) ($0.50) ($0.30) ($0.80)

order/purchase referral upgrade renewal register/apply fees return complaint request change of address site visit

Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

Bank Story Deposits Fees Back office maintenance Teller Transactions ATM Transactions Phone VRU calls Direct Deposits

+ + + + + + -

Customer value

+ ++ ++ ++ ++ +

$850

Customer A

+ ++ ++ ++ ++ +

$850

Customer B Source: “The Evolution of Direct Marketing: Enterprise Customer Management,” Presentation by ChristopherBecke(Darden ’92), May 7, 1999.

Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

Source: “The Evolution of Direct Marketing: Enterprise Customer Management,” Presentation by ChristopherBecke(Darden ’92), May 7, 1999.

Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

4

Deposit revenues

ATM visits

Teller visits

VRU calls

+ ++ ++ ++ ++ +

-

--- --

-

Customer value

Deposit revenues

ATM visits

Teller visits

VRU calls

+ ++ ++ ++ ++ +

-

--- --

-

$850

Customer A

Customer value

$750

Customer A

+ ++ ++ ++ ++ + Customer B

- -- -- - -- -- - ---------- -- -- - --- -62 -- ---- --- -annually --- ---- - -- -- - ------------- - -- -- - - -

- -- -- - -- -- - ---------- -- -- - - ---73 -- ------- -annually ------- - -- -- - ------------- - -- -- - - -

- -- - - -- - - -- - -- --- --- -- - ---------------------- -- ----- -- - ---------130 --- --------------- - ---annually ----- -------- - -- --- --- -- - ---------------------------- - -- -- -- -- -- - - -- - -

$850 Customer B

Source: “The Evolution of Direct Marketing: Enterprise Customer Management,” Presentation by ChristopherBecke(Darden ’92), May 7, 1999.

Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

Deposit revenues

ATM visits

Teller visits

VRU calls

+ ++ ++ ++ ++ +

-

--- --

-

- -- -- - -- -- - ---------- -- -- - --- -62 -- ---- --- -annually --- ---- - -- -- - ------------- - -- -- - - -

+ ++ ++ ++ ++ +

Customer value

- -- -- - -- -- - ---------- -- -- - - ---73 -- ------- -annually ------- - -- -- - ------------- - -- -- - - -

- -- - - -- - - -- - -- --- --- -- - ---------------------- -- ----- -- - ---------130 --- --------------- - ---annually ----- -------- - -- --- --- -- - ---------------------------- - -- -- -- -- -- - - -- - -

($175)

Source: “The Evolution of Direct Marketing: Enterprise Customer Management,” Presentation by ChristopherBecke(Darden ’92), May 7, 1999.

Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

A’s Lo

ok Goo

7% of savings customers generate 55% of profits

d

$750 per customer

$750

Customer A

+ ++ ++ ++ ++ + Customer B

- -- -- - -- -- - ---------- -- -- - --- -62 -- ---- --- -annually --- ---- - -- -- - ------------- - -- -- - - -

- -- -- - -- -- - ---------- -- -- - - ---73 -- ------- -annually ------- - -- -- - ------------- - -- -- - - -

- -- - - -- - - -- - -- --- --- -- - ---------------------- -- ----- -- - ---------130 --- --------------- - ---annually ----- -------- - -- --- --- -- - ---------------------------- - -- -- -- -- -- - - -- - -

($175)

Source: “The Evolution of Direct Marketing: Enterprise Customer Management,” Presentation by ChristopherBecke(Darden ’92), May 7, 1999.

Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

The 80/20 Rule • The top 20% of customers account for 80% of company profits. – Is this surprising? – The fact that 20% account for 80% is not as important as knowing who the 20% are. – Conventional accounting (quarterly company P&L) only tells us the profit from 100% of the customers. Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

B’s Look Ba

10% of checking customers generate losses = 50% of total profits

d

($175) per customer

Source: “The Evolution of Direct Marketing: Enterprise Customer Management,” Presentation by ChristopherBecke(Darden ’92), May 7, 1999.

Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

LifeTime Value (LTV) • That was then, this is now – So far, we have only talked about measuring the past performance of our customer relationships. – While this is helpful, we’d also like to forecast the value of those relationships into the future. – Ultimately, we’d like to figure out what we can do to maximize the value of our relationships. – In particular, we’d like to value proposed new relationships so that we can make intelligent investments in customer acquisition. Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

5

LifeTime Value of a Customer (a.k.a. Customer Equity) • the discounted sum of all future customer revenue streams minus product and servicing costs and remarketing costs.

LifeTime Value of a Customer (a.k.a. Customer Equity) • LTV is the expected NPV of the cash flows from a customer relationship. • Just like we use NPV to evaluate investments and companies, we use LTV to evaluate customer relationships.

Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

Modeling LTV A Simple LTV Model

Estimating LTV • First, find out what LTV is on average. – Calculate LTVs for representative current customers at the date you acquired them. The resulting number is an estimate of the LTVs of customers you are about to acquire. This number is VERY helpful in making prospecting decisions. – Historical versus “Forward-Looking” estimates of LTV? Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

A Simple LTV model EXPECTED CASH FLOWS t=0

$M - $R

t=1

r $M – r $R

t=2

r2 $M – r2 $R

With a little algebra, someone came up with a formula for the NPV of these expected cash flows. LTV = [$M – $R]×[(1+d)/(1+d-r)]

t=3

r3 $M – r3 $R etc.

Click anywhere to see it.

ASSUMPTIONS $M

Contribution per period from active customers.

$R

Retention Spending per period per active customer.

r

retention rate (fraction of current customers retained each period)

d

discount rate per period

Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

Example An Internet Service Provider charges $19.95 per month. Variable costs are about $1.50 per account per month. With marketing spending of $6 per year, their attrition is only 0.5% per month. At a monthly discount rate of 1%, what is the LTV of a customer?

$M = $19.95 - $1.50 = $18.45 $R = $6/12 = $0.5 r

= 0.995

d

= 0.01

LTV = [$M – $R]×[(1+d)/(1+d-r)] LTV = [$18.45 – $0.5]× [1+.01)/(1+.01-0.995)] LTV = [$17.95]× [66.3] LTV = $1,209

Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

6

Lifetime Customer Value Comparison

Future Value Multipliers as a function of discount and retention rates r= d= discount rate retention 0.1 0.11 1.00 11.00 10.09 0.95 7.33 6.94 0.90 5.50 5.29 0.85 4.40 4.27 0.80 3.67 3.58 0.75 3.14 3.08 0.66 2.50 2.47 0.50 1.83 1.82 0.00 1.00 1.00

Assumes 15% discount rate, 4-year customer life for all but Rx which is 7-year customer life. $40

0.12 9.33 6.59 5.09 4.15 3.50 3.03 2.43 1.81 1.00

0.13 8.69 6.28 4.91 4.04 3.42 2.97 2.40 1.79 1.00

0.14 8.14 6.00 4.75 3.93 3.35 2.92 2.38 1.78 1.00

0.15 7.67 5.75 4.60 3.83 3.29 2.88 2.35 1.77 1.00

0.2 6.00 4.80 4.00 3.43 3.00 2.67 2.22 1.71 1.00

For example, a discount of .10 and a retention rate of .95, customer producing $10 in per period contribution after retention spending would have an estimated LTV of $73.30. Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

LTV in Practice But evaluating success solely on the basis of ROI may be taking a nearsighted view of business. That’s why iGo also measures the long-term value (LTV) of its customers, based on their purchases over a period of months or years. It uses an algorithm-- which is based on a present-value annuity model -- that measures customer retention by how often customers visit the website and how much they buy. Investments in customer service, content and advertising are modeled by balancing the anticipated increase in retention against the costs. Recently, for example, iGo looked at LTV data for repeat buyers and first-time website purchasers, and then changed the look and feel of its site, believing that the change would generate more long-term sales.

*Vertically Integrated Specialty *DirectMail Apparel *Dept Store Apparel

Per $30 Order Contribution

*Cash Rx *Off-price Apparel

$20

*3rd Party Rx *Bookstore *Pure -play Apparel

$10

*Grocery

*Drugstore

0 -$250

0

$250 $500 $750 Customer Lifetime Value

$1,000

Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

Estimating LTV • Next, find out how LTV varies by – Source. Are customers who typed in your web address more or less valuable than those referred by Yahoo? – Offer. Are customers acquired through special promotions just as valuable? – Pricing plans. Does the pricing plan the customer chose affect his/her LTV? – Demographics. Are women more valuable than men? – Usage Patterns. Are weekend shoppers more or less valuable than those shopping during the week?

Source: Nick Wreden, Computing Your Website Payoff, by Nick Wreden, Beyond Computing, April 2000, Volume 9, Number 3 School Foundation Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden

The Sports Illustrated Sneaker Phone Promotion HUGE SUCCESS…it brought in lots of new customers and cost per acquired customer (trial) was low.

DISMAL FAILURE…..those customers had very low LTVs, because of low conversion of trial subscriptions.

Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

Acquiring credit card customers using the WWW HUGE SUCCESS…relative response rates and cost per acquired customer were great compared to other media.

DISMAL FAILURE…..However, the acquired customers had above average default rates.

Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

7

The Technology of Tracking

The Importance of Tracking

• Dell print ads contain “e-values” codes. Visitors to the site enter the code…and Dell knows exactly what ad brought them to the site. • Catalog codes and tailored 800 numbers serve the same purpose. • Train your customer service reps to ask, “how did you hear about us?” • Online..do you want to track the site they just came from? the search word they used to get there? • Most importantly…make sure you add the information to your database….it’s almost impossible to retrieve later. Source: Leo Cullum, The New Yorker, 7/94

Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

Estimating LTV • Ultimately, run experiments to learn how to maximize LTVs. – Every contact with your customers is an opportunity to test your customer retention strategies. – For example, if you haven’t heard from someone in six months should you? – – – – –

send a “we miss you” email phone them mail them a coupon. do nothing. do nothing, but if they do return say “we missed you”

– You should…test each of these strategies, track the resulting LTVs and see which one leads to the highest LTV. Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

Example An Internet Service Provider charges $19.95 per month. Variable costs are about $1.50 per account per month. With marketing spending of $6 per year, their attrition is only 0.5% per month. At a monthly discount rate of 1%, what is the LTV of a customer?

$M = $19.95 - $1.50

Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

The Challenges of Testing • The firm has to admit it doesn’t have all the answers. • The firm has to try things it believes aren’t the best. • Tests add work for everyone. (Instead of mastering one way to handle a situation, service reps have to randomly apply k ways.) Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

Example (continued)

$R = $6/12 = $0.5

If the firm cuts retention spending from $6 to $3 per year, they expect attrition will go up to 1% per month.

r

= 0.995

Should they do it?

d

= 0.01

= $18.45

LTV = [$M – $R]×[(1+d)/(1+d-r)]

Don’t move to the next slide until you have an answer.

LTV = [$18.45 – $0.5]× [1+.01)/(1+.01-0.995)] LTV = [$17.95]× [66.3] LTV = $1,209

Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

8

Example (continued) To decide, we need to recalculate LTV under these new assumptions. If the new LTV is higher, we should do it. Otherwise, we should not.

Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

Example (continued) The new LTV is LOWER. The savings in retention spending is NOT worth the increased attrition. The firm should stick with the $6 retention spending.

Example (continued) An Internet Service Provider charges $19.95 per month. Variable costs are about $1.50 per account per month. With marketing spending of $3 per year, their attrition will be 1% per month. At a monthly discount rate of 1%, what is the LTV of a customer?

$M = $19.95 - $1.50 = $18.45 $R = $3/12 = $0.25 r

= 0.99

d

= 0.01

LTV = [$M – $R]×[(1+d)/(1+d-r)] LTV = [$18.45 – $0.25]×[1+.01)/(1+.01-0.99)] LTV = [$18.2 ]× [50.5 ] LTV = $919

Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

Example—acquisition spending This same ISP spends $5 per prospect to initiate the first signup. They are successful 5% of the time. They have considered a more aggressive acquisition plan costing $10 per prospect. What success rate must this new plan have in order to be preferred to the old one? (Assume the ISP must use one plan or the other, and cannot use both.) You are encouraged to try this one on your own. A solution appears on the next slide.

Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

Example—acquisition spending With the old plan, $5 buys them a 0.05 chance of acquiring a new customer. If the new customer has a LTV of $1,209, the expected NPV of the $5 investment is 0.05*$1,209 - $5 or $55.45. In other words, the LTV of a prospect is $55.45. Again, under these assumptions the ISP should be making as many of these investments in acquisition as possible. For the new plan to do just as well, its acquisition rate must be such that the LTV of a prospect is at least $55.45. A little algebra or goal seek shows that the acquisition rate must be at least 0.054. Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

Modeling LTV These examples should give you a taste for the kinds of analysis that are possible using this simplest LTV model. If the firm was able to develop schedules of how r varied with $R and how the acquisition rate varied with acquisition spending, it would be able to find the optimal balance between retention and acquisition spending.

Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

9

Modeling LTV More Complicated Models

Modeling LTV More Complicated Model Consider an online content provider selling annual subscriptions to its service. Regular subscriptions are priced at $99 with a one-year initial trial available for $56.

Only in very simple situations is it possible to write a formula for LTV. When the situation gets more complicated, build an electronic spreadsheet model to calculate LTV.

By company policy, subscribers are not automatically renewed each year. The average amount the company spends in efforts to renew someone’s subscription depends on whether or not they achieve success. The renewal cost for successful renewals averages about $2. For unsuccessful renewals, the cost is $11.

See the next slide for an example.

The incremental cost to provide the content to each subscriber is negligible. Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

Modeling LTV

Modeling LTV

More Complicated Model

More Complicated Model

Historically, the company converts about 40% of its initial $56 trial subscribers into $99 regular subscribers. Renewal rates for regular subscribers are running 70% the first year, 90% the second year, and are expected to level off at 95% each year thereafter. The firm uses a 15% hurdle rate.

SUMMARY OF SITUATION

Fire up your spreadsheet and try this one yourself.

Initial subscription price $56 Conversion rate 40% Regular subscription price $99

Move to the next slide when you are ready to look at a solution.

Renewal rates 70%, 90%, 95% thereafter.

What is the LTV of a new trial subscriber?

Variable Cost = $0

What is the LTV of a first-time regular subscriber?

Renewal Spending = $2 if successful, $11 if not.

What is the LTV of a two-time regular subscriber?

Hurdle rate = 15%

What is the LTV of a three-time regular subscriber?

Horizon—use 15 years.

Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

Modeling LTV

Modeling LTV

More Complicated Model

More Complicated Model

Renewal cost if success $ Renewal cost if fail $ Conversion rate Renewal rates

Hurdle rate Year Number of subscribers Revenue Renewal Costs Cash Flow NPV LTV

$ $ $ $ $

Click here if you want a detailed explanation of this spreadsheet. Click here if you want to move on.

2 11 0.4 0.7 0.9 0.95 0.15 0 1000 56,000 7,400 48,600 186,059 186.06

$ $ $ $ $

1 400 39,600 1,880 37,720 158,078 395.19

$ $ $ $ $

2 280 27,720 812 26,908 138,411 494.33

Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

$ $ $ $ $

3 252 24,948 617 24,331 128,229 508.84

$ $ $ $

4 239.4 23,701 587 23,114 119,483

Renewal cost if success $ Renewal cost if fail $ Conversion rate Renewal rates

Hurdle rate Year Number of subscribers Revenue Renewal Costs Cash Flow NPV LTV

2 11 0.4 0.7 0.9 0.95 0.15

0 1000 $ 56,000 $ $ 7,400 $ $ 48,600 $ $ 186,059 $ $ 186.06 $

We arbitrarily start with 1,000 trial subscribers

1 400 39,600 1,880 37,720 158,078 395.19

$ $ $ $ $

2 280 27,720 812 26,908 138,411 494.33

$ $ $ $ $

3 252 24,948 617 24,331 128,229 508.84

4 239.4 $ 23,701 $ 587 $ 23,114 $ 119,483

Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

10

Modeling LTV

Modeling LTV

More Complicated Model

More Complicated Model

Renewal cost if success $ Renewal cost if fail $ Conversion rate Renewal rates

Hurdle rate Year Number of subscribers Revenue Renewal Costs Cash Flow NPV LTV

$ $ $ $ $

40% convert to regular subscriptions

2 11 0.4 0.7 0.9 0.95 0.15 0 1000 56,000 7,400 48,600 186,059 186.06

Renewal cost if success $ Renewal cost if fail $ Conversion rate Renewal rates

Hurdle rate

$ $ $ $ $

1 400 39,600 1,880 37,720 158,078 395.19

$ $ $ $ $

2 280 27,720 812 26,908 138,411 494.33

$ $ $ $ $

3 252 24,948 617 24,331 128,229 508.84

$ $ $ $

4 239.4 23,701 587 23,114 119,483

Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

Year Number of subscribers Revenue Renewal Costs Cash Flow NPV LTV

70% resubscribe at end of year 2.

2 11 0.4 0.7 0.9 0.95 0.15

0 1000 $ 56,000 $ $ 7,400 $ $ 48,600 $ $ 186,059 $ $ 186.06 $

1 400 39,600 1,880 37,720 158,078 395.19

$ $ $ $ $

2 280 27,720 812 26,908 138,411 494.33

$ $ $ $ $

3 252 24,948 617 24,331 128,229 508.84

Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

Modeling LTV

Modeling LTV

More Complicated Model

More Complicated Model

Renewal cost if success $ Renewal cost if fail $ Conversion rate Renewal rates

Hurdle rate Year Number of subscribers Revenue Renewal Costs Cash Flow NPV LTV

$ $ $ $ $

2 11 0.4 0.7 0.9 0.95 0.15 0 1000 56,000 7,400 48,600 186,059 186.06

$2 per success, $11 per failure

Renewal cost if success $ Renewal cost if fail $ Conversion rate Renewal rates

Hurdle rate

$ $ $ $ $

1 400 39,600 1,880 37,720 158,078 395.19

$ $ $ $ $

2 280 27,720 812 26,908 138,411 494.33

$ $ $ $ $

3 252 24,948 617 24,331 128,229 508.84

$ $ $ $

4 239.4 23,701 587 23,114 119,483

Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

Year Number of subscribers Revenue Renewal Costs Cash Flow NPV LTV

2 11 0.4 0.7 0.9 0.95 0.15

0 1000 $ 56,000 $ $ 7,400 $ $ 48,600 $ $ 186,059 $ $ 186.06 $

NPV at year 0 of year 0 -15 cash flows is $186,059. 1 400 39,600 1,880 37,720 158,078 395.19

$ $ $ $ $

2 280 27,720 812 26,908 138,411 494.33

$ $ $ $ $

3 252 24,948 617 24,331 128,229 508.84

Modeling LTV

More Complicated Model

More Complicated Model

Hurdle rate Year Number of subscribers Revenue Renewal Costs Cash Flow NPV LTV

$ $ $ $ $

2 11 0.4 0.7 0.9 0.95 0.15 0 1000 56,000 7,400 48,600 186,059 186.06

NPV at year 1 of year 1 -15 cash flows.

$ $ $ $ $

1 400 39,600 1,880 37,720 158,078 395.19

$ $ $ $ $

2 280 27,720 812 26,908 138,411 494.33

Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

4 239.4 $ 23,701 $ 587 $ 23,114 $ 119,483

Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

Modeling LTV Renewal cost if success $ Renewal cost if fail $ Conversion rate Renewal rates

4 239.4 $ 23,701 $ 587 $ 23,114 $ 119,483

$ $ $ $ $

3 252 24,948 617 24,331 128,229 508.84

$ $ $ $

4 239.4 23,701 587 23,114 119,483

Renewal cost if success $ Renewal cost if fail $ Conversion rate Renewal rates

Hurdle rate Year Number of subscribers Revenue Renewal Costs Cash Flow NPV LTV

2 11 0.4 0.7 0.9 0.95 0.15

0 1000 $ 56,000 $ $ 7,400 $ $ 48,600 $ $ 186,059 $ $ 186.06 $

NPV at year 0 per year 0 customers.

1 400 39,600 1,880 37,720 158,078 395.19

$ $ $ $ $

2 280 27,720 812 26,908 138,411 494.33

$ $ $ $ $

3 252 24,948 617 24,331 128,229 508.84

4 239.4 $ 23,701 $ 587 $ 23,114 $ 119,483

Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

11

Modeling LTV

Modeling LTV

More Complicated Model

More Complicated Model

Renewal cost if success $ Renewal cost if fail $ Conversion rate Renewal rates

Hurdle rate Year Number of subscribers Revenue Renewal Costs Cash Flow NPV LTV

$ $ $ $ $

So the LTV of an initial subscriber is $186.

2 11 0.4 0.7 0.9 0.95 0.15 0 1000 56,000 7,400 48,600 186,059 186.06

Renewal cost if success $ Renewal cost if fail $ Conversion rate Renewal rates

Hurdle rate

$ $ $ $ $

1 400 39,600 1,880 37,720 158,078 395.19

$ $ $ $ $

2 280 27,720 812 26,908 138,411 494.33

$ $ $ $ $

3 252 24,948 617 24,331 128,229 508.84

$ $ $ $

4 239.4 23,701 587 23,114 119,483

Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

Year Number of subscribers Revenue Renewal Costs Cash Flow NPV LTV

2 11 0.4 0.7 0.9 0.95 0.15

0 1000 $ 56,000 $ $ 7,400 $ $ 48,600 $ $ 186,059 $ $ 186.06 $

The LTV of a new regular subscriber is $395.

1 400 39,600 1,880 37,720 158,078 395.19

$ $ $ $ $

2 280 27,720 812 26,908 138,411 494.33

$ $ $ $ $

3 252 24,948 617 24,331 128,229 508.84

Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

Modeling LTV

Modeling LTV

More Complicated Model

More Complicated Model

Renewal cost if success $ Renewal cost if fail $ Conversion rate Renewal rates

Hurdle rate Year Number of subscribers Revenue Renewal Costs Cash Flow NPV LTV

The LTV of a twotime regular subscriber is $494.

2 11 0.4 0.7 0.9 0.95 0.15

0 1000 $ 56,000 $ $ 7,400 $ $ 48,600 $ $ 186,059 $ $ 186.06 $

Renewal cost if success $ Renewal cost if fail $ Conversion rate Renewal rates

Hurdle rate 1 400 39,600 1,880 37,720 158,078 395.19

$ $ $ $ $

2 280 27,720 812 26,908 138,411 494.33

$ $ $ $ $

3 252 24,948 617 24,331 128,229 508.84

4 239.4 $ 23,701 $ 587 $ 23,114 $ 119,483

Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

Modeling LTV More Complicated LTV Model • Does it make sense to you that – The LTV of a trial subscriber is $186, – The LTV of a first -time regular subscriber Is $395, – The LTV of a two-time regular subscriber is $494, – The LTV of a three-time regular subscriber is $509?

• Can you explain why these LTVs go up? Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

4 239.4 $ 23,701 $ 587 $ 23,114 $ 119,483

Year Number of subscribers Revenue Renewal Costs Cash Flow NPV LTV

2 11 0.4 0.7 0.9 0.95 0.15

0 1000 $ 56,000 $ $ 7,400 $ $ 48,600 $ $ 186,059 $ $ 186.06 $

Click here if you want to see the spreadsheet. Click here if you want to move on. 1 400 39,600 1,880 37,720 158,078 395.19

$ $ $ $ $

2 280 27,720 812 26,908 138,411 494.33

$ $ $ $ $

3 252 24,948 617 24,331 128,229 508.84

4 239.4 $ 23,701 $ 587 $ 23,114 $ 119,483

Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

Caveats of LTV analysis • While it is good that LTV analysis helps the firm focus on serving individual customers, the firm shouldn’t expect their customers to do everything. The customers are not going to – develop the next big thing – warn the firm about impending competitive entry…the firm will only find out after it’s too late--when the customers have left.

• Share of customer (wallet) (requirements) Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

12

Caveats of LTV analysis

Which firm is doing better?

• The LTV of a customer accounts for the cash flows from that customer. • There are other reasons to acquire a customer beyond his/her cash flows: – Acquiring customers can help discourage competition. – Acquiring customers can help secure financing. – Acquiring customers can make it easier to acquire future customers and can raise the LTVs of future acquired customers—(network effects..think AOL, Ebay) (Michael Jordan).

FIRM A Revenue COGS Marketing Profit

1 833.33 708.33 100.00 25.00

2 $ 1,166.67 $ 991.67 $ 150.00 $ 25.00

Cogs/Rev Mkt/Sales ROS

85.0% 12.0% 3.0%

85.0% 12.9% 2.1%

FIRM B Revenue COGS Marketing Profit

1 $ 1,320.43 $ 1,122.37 $ 173.06 $ 25.00

2 $ 1,385.18 $ 1,177.40 $ 182.78 $ 25.00

Cogs/Rev Mkt/Sales ROS

85.0% 13.1% 1.9%

85.0% 13.2% 1.8%

Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

$ $ $ $

Firm B New Customers Total Customers Sales/Customer Mkt/New Cust Churn rate

$ $

1 1.33 3.33 250.00 $ 75.00 $

2 2.00 4.67 250.00 $ 75.00 $ 20%

3 3.07 6.80 250.00 $ 75.00 $ 20%

4 4.77 10.21 250.00 $ 75.00 $ 20%

5 7.50 15.67 250.00 75.00 20%

$ $

1.86 3.86 342.00 $ 93.00 $

1.97 4.05 342.00 $ 93.00 $ 46%

2.09 4.28 342.00 $ 93.00 $ 46%

2.24 4.55 342.00 $ 93.00 $ 46%

2.43 4.88 342.00 93.00 46%

Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

Which firm is doing better? • FIRM B

– Customer LTVs of $123.

– Customer LTVs of $97.

– Customer acquisition costs of $75.

– Customer acquisition costs of $93.

– 15.67 active customers at end of year 5. – Terminal Value = 15.67($123) = $1,931.

– 4.88 active customers at end of year 5. – Terminal Value = 4.88($97) = $472.

Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

85.0% 14.0% 1.0%

85.0% 14.4% 0.6%

3 4 $ 1,463.16 $ 1,557.07 $ 1,243.69 $ 1,323.51 $ 194.47 $ 208.56 $ 25.00 $ 25.00

5 $ 1,670.17 $ 1,419.65 $ 225.53 $ 25.00

85.0% 13.3% 1.7%

85.0% 13.4% 1.6%

85.0% 13.5% 1.5%

Comparing LTVs FIRM A

FIRM B

$M = $250(.15)

$M = $342(.15)

= $37.50

= $51.30

$R = $0

$R = $0

r

= 0.80

r

= 0.54

d

= 0.15

d

= 0.15

LTV = [$M – $R]×[(1+d)/(1+d-r)] LTV = [$37.5 – $0]× [(1+.15)/(1+.15-0.8)]

• FIRM A

85.0% 13.5% 1.5%

5 $ 3,918.67 $ 3,330.87 $ 562.80 $ 25.00

Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

Which firm is doing better? Firm A New Customers Total Customers Sales/Customer Mkt/New Cust Churn rate

3 4 $ 1,700.00 $ 2,553.33 $ 1,445.00 $ 2,170.33 $ 230.00 $ 358.00 $ 25.00 $ 25.00

LTV = [$51.3 – $0]×[(1+.15)/(1+.15-0.54)]

LTV = [$37.5]× [3.29]

LTV = [$51.3]×[1.89]

LTV = $123

LTV = $97

Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

Which firm is doing better?

FIRM A And we realize this only after looking at CUSTOMERS.

Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

13

Additional Reading (required)

Additional Reading (optional) •

• Stewart Alsop. “The Price Tag on You is What Drives Net Stocks.” Fortune (March 1, 1999). Two-page article with quick and dirty speculative calculations for the LTV of AOL and Amazon customers and comparisons to Market Cap per customer.

•

•

Robert C. Blattberg and John Deighton. “Manage Marketing by the Customer Equity Test.” Harvard Business Review(July -Aug 1996). They define customer equity to be the sum of the LTVs of all current customers and present a version of the simple LTV model. F. Robert Dwyer. “Customer Lifetime Valuation to Support Marketing Decision Making.” Journal of Direct Marketing 3(4) Autumn 1989. Award-winning article illustrating the usefulness of LTV calculations. Don Peppers and Martha Rogers. Enterprise One to One (New York: Currency Doubleday, 1997), chapter 2. This chapter captures the main ideas in the presentation.

You should find a link to this article on the course website.

Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

Copyright © 2000. Paul Farris, Phil Pfeifer, and the Darden School Foundation

14