If you're like me, you enjoy curling up in bed each week with a bag of popcorn and watching the latest episode of Shark Tank. Shark Tank has minimal applicability to a day in the life of a tech VC (although Mark Cuban does add a lot of tech VC theory/outlook) but it is instructive for the types of mistakes that entrepreneurs continually make. In this week's episode, Pretty Padded Room was presented, an online marketplace for virtual therapy services started by a young entrepreneur named Bea Arthur. I have to say, I actually think this idea is pretty good. It may be a bit ahead of its time but there are lots of part-time therapists or stay at home moms with MSWs or Doctorates (it was exclusively female focused) who are looking to add some side revenue to their bottom lines. It's the type of business which is at least worth exploring, because it has a strong network effect, and does solve a pain point for busy professionals, or those looking for less expensive mental care. It wouldn't cost a lot of money to get this off the ground because you could implement many "off the rack" products to build out the technical side with minimal overheard.

The problem is that the businesses was presented by one of the least eloquent (or knowledgable) entrepreneurs the show has ever seen. She was unable to answer Mark Cuban's question on the Life Time Value of her customers, which got me thinking that it might make an instructive blog post.

Life Time Value is simply the expected revenue generated by one of your customers over the course of their term purchasing from you. To get the calculation, you would just average out the totality of your customers. For example, if you had 2 customers, one who had spent $10 with you, and another who had spent $20 with you, your customer LTV would be $15.

There are some complications here, however. How do you know a customer will never come back? Alternately, if you've had a customer for three strong months of purchasing, how do you know how long she'll stay. For modeling and investment purposes, you'd find comparable businesses with some records (for example, a typical flash sales customer, or a typical subscriber to a monthly service) and estimate out that way. Your numbers have to be done on a historical level, but you can begin by estimating future projections and refining with each increase in data.

A further refinement of the LTV model is "cohort analysis." Namely that not all customers behave the same. Perhaps customers who joined six months ago spent more because there were fewer competitors discounting their goods. Or maybe macro economic factors have changed and customers who join this month are expected to spend more than those who joined a year ago. All LTV models should break down customers by a cohort class - typically done by month or quarter. The reason for tracking cohort class, as taught by Eric Lefkofsky in his "Building Internet Startups" course at Chicago Booth, is to assess the "Payback" period for any given business or acquisition channel. Eric suggests looking for businesses with a six month payback - that is where the revenue generated by the customer surpasses the cost of acquiring that customer within six months.

The Life Time Value calculation has been the subject of considerable debate and analysis. For more on this topic, you can read Jeremy Liew from Lightspeed's take or the defining article on the subject by Bill Gurley at Benchmark.

Hope this helps and I apologize if it was too basic.

Critical VC Terms - Life Time Value