Monday, April 30, 2007

Can LTV Really Replace Product-Based Metrics?

For some time now, I’ve been touting lifetime value as the magic cure-all for customer management programs. I still believe LTV is the essential foundation but recognize that product-oriented measurement will not simply vanish once LTV appears. As long products are what people purchase, companies will need product managers to nurture them and will create product-level profit statements to judge their performance.

I’m tempted to wave my LTV magic wand and argue that products themselves could be measured on their LTV contribution rather than traditional profit-and-loss. There’s considerable logic to this: if you had a product that was itself profitable but so annoyed customers that they never purchased from you again, wouldn't you want to know about this? The problem is that LTV measures inherently rely on projections, which feel less reliable to many people than numbers that simply record actual transactions. And, let’s face it, identifying the incremental impact of a particular product on a customer’s LTV is quite a challenge—at least one order of magnitude more difficult than measuring the LTV as a whole.

I don’t have a solution. One option is to double-down on LTV, working to develop the incremental impact measures and to make them credible throughout the organization. You would then essentially banish product profitability as a measure—not literally, but by deemphasizing it in reports and compensation programs. This could probably be done with strong management leadership but it would definitely be a strain.

The other choice is to find a way to use product profitability measures so they reinforce rather than conflict with LTV results. The traditional approach to this is transfer pricing, but nobody ever finds that satisfactory. Nor do I see exactly how transfer pricing fits in here. Going back to first principles, the reason we need to look at product performance is we want to incent product managers to build the best possible products and to ensure they are being sold as broadly as possible. Products do, after all, represent a major investment of corporate resources and we want to maximize return on that investment.

But is it possible to define the “best possible products” in any terms other than their impact on lifetime value? I think not. We’ve already seen why looking at product profitability in isolation is likely to be misleading. So it seems committing fully to LTV is the only real choice.

This is a classic dilemma. On the one hand, “failure is not an option”. On the other hand, no matter how hard we try, failure is a real possibility. So while trying very hard to make LTV work, we have to consider alternatives if we can’t come up with an effective approach. This presumably means some modified version of product profitability that incorporates certain LTV concepts. I don't know what that would look like, but will let you know if I come up with any ideas.

No comments: