Customer Lifetime Value (SMALL BOOK 167-177) * Customer lifetime value (CLV), is the net present value of the cash flows attributed to the relationship with a customer. * The use of customer lifetime value as a marketing metric tends to place greater emphasis on customer service and long-term customer satisfaction, rather than on maximizing short-term sales. * Two approaches to CLV: * Disaggregate (“spreadsheet”)– Complex and cumbersome, but allows you to build in any assumptions you want and see how they affect the result. * Aggregate (“analytic”) – Makes (often strong) assumptions, hard to see the mechanism. Easy to remember and compute, allows you to focus on the big factors.
BASIC CLV MODEL: ( a.k.a Margin Multiple ) It reflects the proportion of one’s per-customer profit margin you expect to recoup over time. m: profit margin per period r: retention rate i: discount rate

Key Insights | Assumptions | * Retention and loyalty are key to long-term value. * Other factors (e.g., discount rate) only have an impact when retention is high. All the handwringing about what discount rate to use is irrelevant in many real-world situations. * Margin matters, but it is independent of the other factors. * Customer-based costing is important, and often absent. | * Profit margins are constant * Defection rates are stable over time * Discount rates are stable over time * The customer’s lifetime is infinite | * | Implications | * | * It is low when business is risky (i.e., the discount rate is high) and high when retention is high. * If your churn is too high, it’s hard to make money from your customers. * Conversely, loyal customers provide substantial value. * Retention rate typically has a greater impact than discount rate. In other words, marketing actions have a greater impact on CLV than does “financial engineering”. |

EXAMPLE:

A health club is opening a new location and is considering a promotional campaign to acquire new customers in that market. It has an average profit margin of $50. It calculates that it retains 60% of its customers from one year to the next. The discount rate is assumed to be 12%. In this case: (57.69 > 50) customer is profitable
Extensions

Increasing Margins: * Suppose we believe that profit margins increase over the life of the customer (or over time). If they increase at a constant growth rate, g, we have: CLV increases exponentially, but the improvement is only substantial for high retention rates and requires high growth rates.

Notes:
It is unrealistic to assume that margins will keep growing at such a high rate indefinitely.
One approach is to assume a shorter time horizon for your customers, if you think that the margin growth process is the same for all customers regardless of cohort. Another possibility is that growth is not a function of the maturity of the customer but rather the firm. In this case, we might expect margins to grow rapidly at first and then to decline in later years.

(Another extension is margins grow at a decreasing rate. This is not mentioned in the study guide and is more complex. Equation. Growth slows at rate (k))

Improving Retention: * We can do something quite similar to model increasing retention. Suppose k reflects the rate of change in retention: rt= r0+ (r – r0)[1-e-kt]

* To determine CLV here, we solve the above equation to obtain the retention rate in each time period. We then compute the fraction of customers remaining at the end of that year: Ft= rtFt-1

* Then we compute the present value of the margin for that fraction of customers using a $1 margin: PVt= Ft/(1+i)t+1

* Finally, we cumulate across t to get the margin multiple.

Implication: Dramatic differences in the speed with which retention rates change have relatively small effects on the margin multiple.

Finite Time Horizons:
We calculated CLV using an infinite time horizon. There are instances where finite time horizons may be more appropriate. To compute CLV over n time periods, we simply start with our original equation with n fixed periods and transform it to: (BELOW)

Implication: Projection horizon can make a big difference for high retention rates or for short horizons. Low retention rates bottom out quickly.

All prior examples compute CLV as cumulative, time-discounted average of the profits that consumers will generate. However some customers won’t make a profit.

Industries where one group of consumers subsidizes another include: auction houses, employment and dating services, real estate brokerages, and shopping malls. (Think of company hosting market place) NETWORK EFFECTS

* Gupta, Mela, and Vidal-Sanz (2006) present a network model that incorporated both direct and indirect network effects. Auctions.com had built separate models to predict future growth among their buyers and sellers. They predicted that both groups would grow rapidly at first, but then would slow, saturating at around 150 months (@2M sellers and 10M buyers). * The Gupta et al. model showed that buyers acquired early were very valuable because they would bring in more buyers and sellers. However, as the business matured, the buyers’ network effect would decrease, resulting in lower buyer CLV.

* Auctions.com considered three strategies: 1. A skimming strategy: charge a high fee for sellers initially to increase short-term revenues, then lower the fee to fend off competition. 2. A penetration strategy: charge a low fee for sellers initially to increase their number then raise rates over time. 3. A constant strategy: Charge the same fee to all sellers.

RESULTS:
Implication: When network effects are strong, they can have dramatic effects on CLV and the profitability of different marketing strategies.

Implication: When network effects are strong, they can have dramatic effects on CLV and the profitability of different marketing strategies.

KEY POINTS * CLV is based on defection rates and the pattern of (discounted) profits over time. * You can get a reasonable approximation of CLV using relatively simple models. * Like all models, these approaches involve assumptions. Luckily, changes in these assumptions generally has minimal impact on CLV, at least in many typical scenarios. * In contrast, network effects can have a large impact on CLV. If you are in an industry where network effects are prevalent, you would be wise to incorporate them carefully.

I DIDN’T INCLUDE THE PARTS ABOUT HOW INTEREST RATES AND RETENTION RATES COMPOUND OVER TIME, FIGURED EVERYONE UNDERSTOOD THEM