There are many different approaches to project prioritization, but the most popular ones are the financial method and the scoring model. In this posting let us examine the financial methodology. In a nutshell it implies choosing some kind of a financial criterion – be it a net present value, internal rate of return or some other formula – and calculating a value for each project. Once the ROI for each project has been calculated, the projects are ranked according to their ROIs in the descending order.
Let us look at an example of how it may happen. We have a company that wants to implement 10 projects and has 200 man-months in their resource pool (roughly speaking 20 people working together for one year including vacation time, and allowances for sick days, etc.)
The list of projects together with their expected ROIs is presented in Table 1:
Next, the company needs to estimated the efforts required for each project and rank the projects according to their ROIs (see Table 2):
It is clear from Table 2 that the company in question can do projects H, E, A, F, C, I and G assuming their projections regarding the projects’ ROIs and efforts required were correct. Adding Project B to the mix will force the company to exceed their effort threshold.
While the purely financial models are very good at instilling the sense of discipline and accountability they all suffer from a couple of inherent problems. One can argue that every financial formula out there can be presented in the following form:
Financial value = f(Revenues/Costs)
In other words any financial value is positively correlated with the expected cash inflows from the project and negatively correlated with the cost of the project.
Numerous studies confirm that our ability to predict project cost at the project inception is somewhere between +300% and -75% for high-risk industries and between +75% and -25% for familiar endeavors (see Figure 1).
On the other hand in many instances the revenue forecasts are even less accurate with a potential array of values anywhere from -100% to +∞.
Let me prove this controversial point by using a couple of examples. When Segway was launched in 2001 it was advertised as the most revolutionary contraption since the invention of the personal computers. The company was forecasting sales of 50,000 units annually. However the company was only able to sell only 6,000 vehicles.
On the other hand, I seriously doubt that when the first iPhone project was being conceived in the mind of Steve Jobs, he expected the sales of this product to exceed the entire annual revenues of Microsoft.
So mathematically speaking we have a fraction in which the numerator can be predicted with an accuracy of +300%, -75% and the numerator with an accuracy of -100%, +infinity. How reliable then is the overall formula?
Another problem with the purely financial models is that they ignore such factors as strategic alignment, fit to the existing supply chain, strategic value of the projects proposed and other important aspects.