Pricing for Service / Maintenance – Understanding the Radioactive Isotope Decay Dynamics could be the key
Yes, if you could predict the probability of number of radioactive decays of Plutonium-239 in a 3 second frame, you could very well effectively Price some of the service / maintenance lines for your company
One of the tricky issues many companies face is how to price the service / maintenance line of their portfolio, for many service function is a cost and deployed to garner customer loyalty , goodwill , subsumed as part of their capital warrant cost or often outsourced.
Due to this approach , companies tend to give it free, or typically underprice – leading to increased cost, inability to monetize this offering and leaving money on the table
The reason why they don’t price or under-price is because most companies are unable to forecast the service instances/ Breakdown at a particular time and therefore end up serving these complaints on ad-hoc basis and hence the unwillingless to price it right
The key starting elements for this could be :
A) Understand from past data – what the trends ( % Breakdown / Average Cost to manage breakdown/ Aging analysis)
B) Model creation – To predict the probability of a given number of service / maintenance instances in a fixed interval of time.
C) Using the above data – building a standard profit – loss model which will inform the bare minimum price
The most critical step in my opinion is get the Step B right, how do we predict the probability of 2021 service requirement using past data – Lo’ the answer lies in the probability of number of radioactive decays of Plutonium-239 in a 3 second frame
Radioactive Decays or any independent events which occur at a constant rate within a given interval of time follow a Poisson Distribution.
Simply put – Let us assume the company brings back a sample of 20 machines and checks what was the defect issues in these machines , suppose it gets the following data
Then we can predict the probability for the entire capital footprint using the Poisson Distribution Methodology , the probability and the average cost will be as below
The minimum price that the company could need to maintain 1 machine is 17,496 $$ and the probability of 7 defects is less than 1 % . This number is very critical for the finance team then to build the cost structures for maintenance and subsequently build the P&L model
If you have an exact probability for breakdown you could also use an binomial model to predict, either way it is critical to forecast the overall .
References :
https://www.itl.nist.gov/div898/handbook/eda/section3/eda366j.htm
https://towardsdatascience.com/the-poisson-distribution-and-poisson-process-explained-4e2cb17d459
Business Consultant and Private Equity Investment
4 年Nice article Anantha, good to see you are continuing to promote best practice pricing strategy ??
GM, Vision Care at Johnson & Johnson
4 年A good pricing methodology could take into account at least the basic underlying price drivers: frequency or arrival of failure, their recurrence over time, types of failure and associated costs. A more detailed approach could integrate customer information, e.g. industry sector and Place of residence (accessibility to service team), and machine characteristics, e.g. usage (no. of cases etc). Using such risk factors then allows for machine and customer specific pricing based on a proper risk assessment. A pricing strategy driven by data analytics may be inspired by the practice of insurance pricing, where predictive models are extensively used to analyze the number of claims (frequency) and their corresponding impact or cost (severity), in the presence of risk factors. Potentially, at least three approaches could be used to price such a contract: first, the price can be based on the predicted costs under the contract; second, it can be based on the perceived value of the full-service contract; third, it can be benchmarked against the price of similar contracts offered by competitors.... There is lot that can be done from a marketing stand point in the second "perceived value" approach :)
Asst Vice President : Product & Business @ Raytex IT Services | HealthTech | MedTech | ServiceNow
4 年Very nice, thanks for sharing this......
Regional Senior Director, Commercial Operations and Strategy at Johnson & Johnson Vision
4 年Very well articulated Dr. Anantha .. depending on velocity ML could play vital role in B. The key here is predictive maintenance (customer delight) and IoT is another game changer there.