Predicting project outcomes is important!

The recent cancellation of the 2026 Commonwealth Games by the Victorian Government is a dramatic example of using predicted project outcomes to minimize damage to an organisation. The escalation in the predicted cost of delivery from $2.6 billion to above $6 billion suggests the original bid was wildly optimistic and discovering how such an error occurred should be worthy of enquiry, but that is not the focus of this post.

Once the predicted costs moved to a point where there was no benefit in continuing with the project, it was terminated. While the cancellation could have been done earlier and far more elegantly the fact remains cancelling the project was by far the best decision.

However, to be able to make this type of call, management need information they can rely on. Even then the decision is not simple.

Cost considerations

A decision to cancel a project has to balance: sunk costs, the cost to complete, the expected benefits, and the cost of not completing the project.  The three elements that matter are the cost to complete vs the benefits (sunk costs are lost either way), and the costs of not completing the project.

For more on sunk costs see: https://mosaicprojects.com.au/Mag_Articles/P022_Sunk_Costs.pdf

Time considerations

Time is usually a secondary consideration but can be vital – the Commonwealth Games facilities would need to be open before the games start!  For more normal projects knowing the current projected completion date and the variance at completion are still important for two reasons.

First, the cost of time matters and needs to be included in the cost to complete estimate. Delayed completion may also impact benefits.

Second and more important, time issues tend to emerge as problem well before cost issues show up. A project that is losing time and is expected to finish late will almost inevitably show negative cost variances sooner or later.  Conversely, fixing the root cause of the time issues will often have a positive effect on the overall costs as well.

The problem is most projects do not run systems that are capable of producing a reliable prediction of the expected completion date. Our recent paper Calculating Completion looked at seven different methodologies for managing a project, only two gave reliable predictions of the completion date; these were Earned Schedule and Work Performance Management.

Earned Schedule (ES) was the best option, but implementing ES requires a significant investment in skills and systems. 

Work Performance Management (WPM) achieved similar results to ES without the overhead. All that is required to use the WPM spreadsheet is three bits of information.  To set up the WPM model the amount of work to be accomplished and the time allowed is needed, any metric can be used provided it is applied consistently. This baseline gives you the amount of work expected to be accomplished by a given date. The other bit of information is the actual amount of work achieved by the date.  From this data the predicted completion date is calculated.

The assumption built into WPM is that work will continue at the current rate. If the result is not acceptable, management needs to do something to change the rate of working. If this is not feasible, then the viability of the project needs to be considered and/or the baseline reset to what is achievable. For more on WPM see: https://mosaicprojects.com.au/PMKI-SCH-041.php#Overview