The One Number To Predict Revenue (Part 4)

It was then that I re-read an article by Daniel McCarthy and Peter Fader in the January-February 2020 issue of Harvard Business Review about Customer-based Corporate Valuation (CBCV). This excerpt stood out:

Recognizing that every dollar of revenue comes from a customer who makes a purchase, CBCV exploits basic accounting principles to make revenue projections from the bottom up instead of from the top down. Although this may seem like a radical departure from traditional frameworks, that’s not the case: CBCV simply brings more focus to how individual customer behavior drives the top line.

What do we need to implement CBCV? In addition to the usual financial statement data, two things are required: a model for customer behavior (what we call the customer-base model), and customer data that we feed into it. The model consists of four interlocking submodels governing how each customer of a firm will behave. They are:

  1. the customer acquisition model, which forecasts the inflow of new customers
  2. the customer retention model, which forecasts how long customers will remain active
  3. the purchase model, which forecasts how frequently customers will transact with a firm
  4. the basket-size model, which forecasts how much customers spend per purchase

Bringing these models together enables us to understand the critical behaviors of every customer at a firm—who will be acquired when, how much they’ll spend over time, and so on. Summing up all the projected spends across customers gives us our quarterly revenue forecasts. Together, these models can produce much more precise estimates of future revenues streams—and from that, one can make much better estimates of what a company is really worth.

As outlined by McCarthy and Fader, CBCV is an excellent metric for measuring performance of public companies who can be expected to disclose the data. What I needed was similar but also simpler – which CEOs could easily track. And that is where the idea of Net Predicted Revenue (NPR) came up – a way to estimate the revenue from all future customers, factoring in churn and new acquisition. Could NPR do for a business what NPS did for customer loyalty? Could NPR be the one number that “solved marketing” the way Jim Simons of Renaissance Technologies solved the market?

Tomorrow: The One Number To Predict Revenue (Part 5)