Blog >
New and Improved CLV2.0 – Generating Demand


New and Improved CLV2.0 – Generating Demand

Jayanth Krishnan shares a company article on a company on an improved metric, inspired by Jobs Theory, that addresses CLV’s limitations in addition to making growth and margin the lingua-franca of the whole organization.


CLV is a sound quantitative metric provided the data fed to it has a sound basis in customer’s fundamental approach to making choices. CLV2.0 is a revised metric inspired by Jobs theory, that addresses CLV’s limitations in addition to making growth and margin the lingua-franca of the whole organization. 

Recap of the problem

CLV does not help you design or build a decent product. It’s great for the nerds to crunch numbers and for C-suite to develop strategy. JTBD thinking is great to figure out customer needs, but it does not tell me which customer is valuable to the business. 

We did the research

We went deep and researched the foundations of JTBD. We studied the behavioral psychology assumptions embedded in the theory. We studied the foundational ideas embedded in demand influencing theories and other alphabet soup theories surrounding the simple (but elegant idea) of Jobs. 

We went meta. We looked at the philosophy of CLV, the different research streams into CLV including marketing science, finance, management science and what management strategy has to say on the topic. We realized that the foundations of customer lifetime value have long been forgotten in the frenzy that has followed big data/data science and computational prowess of contemporary Machine Learning (ML) methods. 

CLV is a fantastic metric to gauge the customer population, only if the data fed into the CLV engine meets some common sense criteria, namely

The data you have on customers is indeed complete i.e. you did not leave potential customers out of your data collection process

The data should reveal customer preferences / needs, not just take customer opinion about solutions/products as reflective of their underlying needs.  In nerd speak, conditional variance in data would lead you to bias your conclusions on CLV. 

These two observations are not just marketing / data fixes. It has huge ramifications to how CLV centric organizations function. When the data fed to the CLV engine is biased, you end up doing 4 things inadvertently

You are orienting your whole organization around an assumption of who your customer is instead of having a fact-base on your customers. That’s a lot of manpower working to serve a subset of customers when you can be serving more customers (at a lower cost). 

Your product, UX/UI and operations are working their tail off and not seeing the results in the numbers. Why? Simple, customers’ actual behavior is not what they claimed it would be in the surveys. Remember, your data collection process? How can UX/UI be blamed when survey data design was flawed to begin with. 


Key points include:

  • JTBD thinking
  • Contemporary Machine Learning methods
  • Customer journey


Read the full article, CLV2.0 – Generating demand, not just estimating demand, on