Wednesday, 27 May 2009

Resource allocation: in series or in parallel?

I remember being taught in undergrad finance that the enterprise "can always find enough resource to pursue all opportunities whose return exceeds the risk adjusted cost of capital" (or words to that effect).   A fine theory. But every organisation I've ever worked in always seemed to be short of resources somewhere.

So, scarce resources need to be allocated.

Should you fully resource you highest priority project, then the next, then the next (i.e. in series)? Or should you spread your resource amongst all of the projects (i.e. in parallel)?

This can be visualised as shown below.   In the top half of the diagram we see resources allocated in series - only once Priority 1 has all of the resources it needs, do resources overflow into priority 2.   In this case, priority 4 gets nothing.   In the bottom half of the diagram we see resources allocated in parallel based on the weighting, or relative importance of each of the projects.   (In practice, of course, the buckets would also be of different sizes.)
Most organisations I have encountered have a tendency to resource projects in parralel.   Everyone gets a little resource to appease their demands.   It's a political thing as much as anything else.

But there are three good reasons why it's better to allocate resources to projects in series:
  1. From a practical perspective, it makes sense to allocate resources in series.   It's better to have one project properly resourced and with a chance of success, than it is to have two projects inadequately resourced and with less chance of success.
  2. In a pure financial sense, it also makes more sense to resource projects in series.   Consider these respective NPV calculations (discounted at 15%) where project A and project B both cost £200 and have a payoff of £300.
  3. In seriesNPVYear 1Year 2Year 3Year 4Year 5
    Project A34.68-100-10030000
    Project B26.2300-100-100300
    TOTAL60.91

    In parallelNPVYear 1Year 2Year 3Year 4Year 5
    Project A6.40-50-50-50-50300
    Project B6.40-50-50-50-50300
    TOTAL12.80

    The NPV from funding the projects in series is clearly higher.   Not to mention the additional option value of being able to decide not to start Project B at all at the later date.

  4. It's better for morale - when the first prject finishes, it will give people a morale boost which will carry forward to the later project.

That's not to say you'll never run projects in paralel, just that you should apply all of the resources project A can usefully use before you look at what is available for project B.

For example, not all resources are created equal, and those not suitable to project A might be usefully applied to project B before project A completes. (Consider for example a Marketing intensive project A with few IT implications and an IT intensive project B with few marketing implications.)

So, in practice, whilst there will always be some parallelism in projects, serial resource allocation should be you starting point.