In a 60-ball, “pick 6 balls to win” lottery, there are over 50 million possible outcomes. In a 40 project portfolio, there are over 1 trillion possible subsets of projects for you to choose from.
And the number of possible subsets grows exponentially as the project portfolio grows. So if you add just 5 more projects to your 40 project portfolio, you now have over 35 trillion possible portfolio subsets to choose from.
So how do you find the set of projects that delivers the most value to your firm while also not exceeding your capital, resource, timing, and risk constraints?
If you’re using a spreadsheet and trying to do it manually, the odds of finding an optimal portfolio are far, far lower than winning the lottery. This is the reason why real project portfolio optimization is so important to project portfolio management for getting the maximum value from your project portfolio.
There is a lot of misunderstanding about what project portfolio optimization actually is. It isn’t about so-called balancing a portfolio or just picking the most valuable projects or simply reaching an agreement about a certain set of projects to execute. When I talk about real optimization, I’m speaking to extracting the maximum value from your project portfolio within the limits of multiple simultaneous constraints such as limited capital, limited resources, limited time, and tolerance for risk.
So real optimization includes simultaneously:
- Maximizing value from constrained capital
- Maximizing value from constrained resources
- Maximizing value from time constraints
- Controlling risk
- Including non-financial and strategic project value
- Managing uncertainty
- Including for non-discretionary spending
- Accounting for complex project dependencies
This is the real problem that portfolio optimization solves. And it basically involves 2 steps:
1: Prioritizing your projects by a value score, and
2: Using an optimizer to maximize the overall value of the selected portfolio while not exceeding your constraint limitations.
Now, in regards to prioritizing projects, I have written about that in other Ezine articles, but I want to mention here that a solid, defensible, and robust prioritization methodology is essential for project portfolio optimization.
Because if you’re trying to maximize project portfolio value but the value scores for individual projects aren’t defensible and robust, then your optimization is going to be more motion design portfolio or less worthless because it is based on essentially erroneous value scores.
So let’s assume that you’ve got a good set of values scores for your projects and you’re ready to optimize your portfolio. How do you do this?
The first thing you want to do is to understand what your constraints are. We have talked about some of the obvious ones already-things like capital costs, resources, and risk, but there are others such as project dependencies, timing, and non-discretionary projects. All of these constraints need to be considered simultaneously when you’re optimizing your portfolio.
If you try to do this with a spreadsheet you’ll immediately see the problem.
If you just pick the highest value projects by going down the list until you run out of capital, you may find that you don’t have enough resources to do all those projects.
So then you might find a set of projects that meets your cost and resource constraints only to find out that the set is too risky. So then you work hard trying balance your cost, resources, and risk constraints only to find out that it doesn’t meet your project dependencies constraint.
And so you keep trying to do this manually on and on and on, and even if you finally find a solution that meets all the constraint requirements, you still won’t know whether or not it is an optimal or near optimal solution and based on the sheer numbers of possibilities the chances are that it isn’t.
Which is why it is critical to use an optimizer to help you find optimal and near optimal portfolio subsets in order to extract the most value from your project portfolios. A good portfolio optimizer will find optimal or near optimal solutions, where a near optimal solution is one that is within a few percentage points of the absolute optimal solution.