Project selection criteria – are you doing the right projects?

There’s a lot of attention, and quite correctly in my view, paid to making sure that projects are done right. And by that I mean using the appropriate methods, resources, reporting templates, etc.  But how much effort is spent on project selection criteria – or making sure you’re doing the right projects?

There are always some projects that are ‘no brainers’; regulatory, essential maintenance, previously announced, and so on. But once these have been allowed for, and assuming there is people, time and money left over, which projects should we do?

The answer is, fundamentally, very simple – the ones that add the most value.  But can we use that as a means of creating project selection criteria to select the ‘right’ projects?

It’s easy enough to say that we should go for the ones that will give us the ‘greatest value’, but what does that mean? Something that might be valuable to one person, or organisation, might not be to another, and to make things even more complicated, things change over time.  We might also discover that all the high-value projects use the same specialist resources, so doing them all isn’t an option anyway.

To illustrate the point, suppose we have a choice of running one or other of the projects in the table below:

If the economy is roaring like a lion, your project selection criteria might well pick Project Lion, but in a financial crisis there are more project Tigers around.  There are other things to consider, too.  It is generally true that cost avoidance benefits are less risky than increased sales benefits. For example, avoiding the need for software licences does not rely on the fickle consumer like the increased sales do.

As we’re considering the risk to benefits, we should also be considering the risk to costs – how confident are we in the budget and timescale?  We will also have to find a way of comparing financial benefits with strategic benefits, and take account of resource constraints as well as the financial ones that obliged us to pick between Projects Lion and Tiger in the first place.

If we can do all these things, so that all projects can be assigned values for cost, benefit and use of scarce resources, then we ought to be able to work out the costs and benefits of all of the possible project portfolios. That will then allow us to pick the portfolio with the most benefit for the cost involved.  Oh, and we need to allow for dependencies between projects, and also take account of synergies and avoid double-counting of benefits.  Easy!

Except, of course, it isn’t.  If we have eight candidate projects, there are 256 possible portfolios – do none, do them all, and every option in between.  We could sort them out on a spreadsheet in an afternoon.  However, if there are 28 projects, then there are over 256 million possible portfolios. Imagine how long that would take!  (Nearly a year at ten per second, 20 hours a day with no weekends off…)  You might need some help with this.