Utility based pricing: not the holy grail in price models
The price model is an important factor in outsourcing contracts. The overall trend in price models is towards utility based pricing where standardized services have a fixed price. Then the total price of a set of services is calculated by P(rice) times Q(uantity). Typically we refer to this model as PxQ. It greatly simplifies comparing multiple vendors in a tender situation which explains the wide-spread use in selection processes. It may however not be the best model for running an engagement.
At first glance PxQ has the beauty of simplicity and it addresses the usual requirements clients have towards price models. It is predictable, transparent and it provides cost flexibility. By no means however does it reflect the cost structure of a supplier. For that reason parties often agree on additional constraints. The model is only applicable within certain bandwidths and even then volume discounts are applicable. Even worse, they add a whole set of penalties (effectively taking quality into the equation) and incentive schemes. Simplicity is then the first victim. In scenario’s where services are not completely standardized the conservation law of complexity applies. Reducing complexity in the price model will increase the complexity in the description of the services.
In some cases PxQ actually hurts
For instance when drafting a price model for application services one can strive for utility based pricing. In many of these attempts I have witnessed and engaged in wide academic discussions. Many times I contemplated the feasibility of Functions Points, Story Points and Complexity Points with clients and vendors. Obviously the applicable development methods (Waterfall vs. Agile), technology, the stage of applications in their life cycle and mutual dependencies between client and vendor complicated the models further. Flexible gearing factors, average resource pyramids with changing offshore ratio’s and productivity targets improving over time added even more complexity. In all of these cases the original intention was to create an easy PxQ-model for the engagement.
When you require significant additional complexity to fit a calculation into an utility based model, chances are PxQ is not the correct model. In those cases parties should go back to the drawing board and re-visit their starting points.
For commodity services PxQ is an excellent model. But that does not make utility based pricing a holy grail for every type of engagement. Especially in contracting application services many assumptions, averages of disparate service components and complex calculations are needed to make services appear countable. When contract negotiations spin out off control in ever larger complexity of service descriptions, letting go of the simplest of all price models may be the lesser of many evils. I highly appreciate any feedback on candidates for the all-but-one simplest price model for application services!
