How Do YOU Measure Customer Lifetime Value?

In the October 2007 issue of Harvard Business Review, an article titled How Valuable Is Word Of Mouth discusses the distinction between customers’ lifetime value and referral value. The article say this about CLV:

Estimating a CLV is relatively straightforward. The value to FirmCo of all that Mary will ever buy equals the amount that her purchases will contribute to FirmCo’s operating margin minus the costs of marketing to her.”

Oh really? And what should we do with the costs of providing service to Mary during her tenure as a customer? Ignore them? Assume they’re equal across customers?

The article goes on to say:

No one really knows how much Mary will buy from FirmCo in the future, but we can make an estimate by analyzing her past purchases over some period of time…then projecting that pattern forward using sophisticated statistical models.”

This is a fairly common practice in many firms. This approach ignores two important factors, however:

1) Life stage events. In research I did that looked at the impact of moving on consumers’ purchase habits, I found — not surprisingly — that consumers who move make a lot of purchases in and around the time of their move. But, more interestingly, many consumers change their ongoing purchase habits — sometimes spending more, sometimes spending less — in many product/service categories, depending on the reasons for the move. Most CLV calculations, even those employing “sophisticated” models, miss these events.

2) Moments of truth. A term coined by McKinsey, these customer interactions leave an indelible mark on the relationship. If positive, they can amplify the relationship — if negative, they could kill it altogether. I’m not aware of any firm that explicitly incorporates the possibility or likelihood of these “relationship disrupters” in their CLV calculations.

The HBR article makes a case for why — and how — to incorporate referral value into a CLV calculation. But marketers should also: 1) incorporate service costs; 2) account for life stage events; and 3) model for relationship disrupters.

Technorati Tags: , ,


7 thoughts on “How Do YOU Measure Customer Lifetime Value?

  1. Here are the conclusions from the HBR article that caught my attention:

    #1 Most good intentions remain just that – good intentions.
    Only a third of people who said they would refer someone actually did. What’s more, very few referrals actually led to new customers. Of those who became customers, only about 10% became profitable.

    #2 Your most loyal purchasers are not your most valuable marketers. It turns out that the customers who buy the most from you are not your best marketers. What’s more, your best marketers may be worth far more to your company than your most avid consumers.

    #3 Half of customers’ referrals are made to those who would have become customers anyway, whether they received a referral or not.

    #4 Referrals made by customers following an incentive campaign can be attributed to that campaign for about a year.

    I think Ron’s more critical of the study than me. Most studies will have oversights and variables that are difficult (if not impossible) to predict or quantify. But one thing is for sure: This study goes wayyyy beyond the oversimplified NPS model. I appreciated what the study attempts to accomplish – warts and all.

  2. JP — Thanks for the comment. I’m really not critical of the study. In fact, if anything, it supports my criticisms of simply measuring referral intention. What I was calling it out on was the CLV calculation.

  3. I have to ring in here..

    This study is not the same, better or challenging Net Promoter Score in my opinion. Because the study uses bribes…er incentives.

    Reichheld specifically states in his book The Ultimate Question that you cannot pay, coerce or give a chance to win something to get a person to complete the Net Promoter Score. Taints the data.

    We’re not even comparing apples to apples here guys. Leave NPS alone!!

  4. It’s also prudent to use modeling techniques like RFM (Recency, Frequency and Monetary Value) to help determine Lifetime Value and discover who your best customers truly are based on facts.This will help determine who you should target in the future.

    Life event or life stage marketing works because, if done correctly, it deals with relevancy — the key driver to marketing success. Graduations, weddings, births, marriages, divorces, deaths occur daily and impact lives, so marketers need to use these events to be relevant.

    As for customer service, it should be the FIRST thing every company addresses. Poor customer service is one of the fastest ways to lose customers and shutter the shop. Great customer service alone can contribute to significant growth.

    Good post. Thank you.

  5. Ron, you point out that the HBS article’s definition of CLV is missing the cost to serve the customer. I agree. It’s also missing the value of influence. Word of mouth is not new and social computing hasn’t suddenly created a whole new dimension to the question of customer value. As early as 1995 when I was with Tessera Enterprise Systems we were working with clients to incorporate influence into CLV calculations. In 2000, our former colleague, Erin Kinikin, wrote a report that provided a general equation for CLV:

    Customer LTV = Value (Initial revenue – Costs) + NPV(Loyalty*(Future Revenue – Costs)) + NPV(Loyalty*Influence Value)

    (NPV is Net Present Value, the best proxy for Loyalty is retention (i.e., avg years that the customer will remail a customer)

    As you see, costs and influence are crucial components of the calculation. Of course, the challenge with a metric like this is that it is really hard to calculate. Few firms really understand things like cost to serve at an individual customer level. So, the best most can do is to use estimates and proxies. But, while hard, I it provides a heck of a lot more richness than metrics like NPS.

    The beauty of NPS is that it’s so darn simple. But it causes me to wonder what the NPS score of those two old grannies that were “fired” as customers by Filene’s Basement a couple of years ago. The story was that they several afternoons a week shopping at Filene’s Basement. The problem was that they also quibbled with service staff over every pull to get another 10% off and returned a ton of the merchandise that they bought later. Filene’s “fired” them because they were so unprofitable (and unproductive) to their business. Before they were “let go”, I bet these ladies probably told their friends about the all the great deals they were getting — thereby sending in more customers just like them.

  6. Calculating the lifetime value of a customer indeed it is important. But as I do notice that currently most of the young people were shifting to another bank’s due to certain reasons ( changing of work or a better rates from other banks). And I am looking forward of any article that are discussing about this issue.

  7. Just my two cents here… The costs to service Mary are taken into account in the HBR article. The calculation is not the amount of purchases from Mary, but rather her contribution. Contribution (aka – net marketing contribution) is basically revenue minus variable costs (which will include the cost of service).

Comments are closed.