Project Managers Risk Averse? - Part II (Dec 24, 2010)

A recent (late 2010) “Linkedin” discussion started with: “I was at a live project manager networking event the other night and someone said, ‘Project Managers are Risk adverse by nature.’ Is risk adversion a quality of a successful project manager?” What are they really talking about? The discussion seems to conclude project managers are not anymore risk adverse or averse than any other person. What they do is manage their risk. I thought I would write this article about risk in projects, how risk is planned for (in the project plan) and how risk is managed during project implementation.

2. Risk and Project Planning

Is it safe to assume that project managers are no different than any other manager – they want to have a good career and get ahead in life? Are the three key project deliverables scope, budget and the project due date? If the answer to both questions is YES, then it follows that (almost) every project manager wants to be reliable. He wants to deliver the required scope within the budget on or before the due date.
To ensure reliability has to begin during planning. The plan is the best possible representation of what will eventually happen in reality – or that is what it should be. The project manager must consider all the tasks necessary to complete the project, together with the risks he may well encounter. Not to do so will almost certainly ensure his own and his project’s failure. The experienced project manager knows there are many risks involved with every project – here is an incomplete list:

1. He might miss some important activities that should be included in the project.
2. The client may modify (increase) the project’s scope during execution.
3. There are (many) other projects underway that require the same resources as his project.
4. Project priorities are unclear and are often changed.
5. People (resources) will not always be available when needed (or when promised).
6. There are always surprises with planned activities – they can easily take significantly longer.
7. Authorisations (paperwork) may not be delivered when planned.

Uncertainties may or may not happen – nevertheless they must be considered in the plan. The project manager may or may not be expert in the various activities in a project – often he must rely on the time estimate of the individual (the expert) that will execute an activity. The people that perform an activity have the same goal as their project manager(s), they also want and need to be reliable. The time and cost it will take to execute and deliver (all) the scope determines the value (return on investment) of a project. There is very likely to be pressure on both budget and project time (projects that are too expensive or take too long will not be economical and are likely to be dropped). While management will put pressure on budget and timing, they too need reliability. Too much pressure will jeopardize reliability and their credibility. Project planning seems to be a very difficult task – fraught with risk!
What is the shape of risk? For every task there is some sort of minimum time – the task cannot ever be completed more quickly than that. There is also a most likely time around which the task would be finished most of the time (would we be able to execute it many times). There is no maximum – a big enough disaster, while unlikely, can occur so that a task is never finished (think of the roof of the Montreal Olympic stadium – it was completed something like 16 years after the Olympics there). The probability curve of a task is a skewed curve – with a long tail to the right. The situation looks like the graphic above.

The importance of the curve is that it shows how much safety a resource must build into his time estimate in order to be 85% certain of finishing on time. (The curve will always have

this general shape though the amount of skew to the right will vary.)

Each task is estimated so that the resource concerned (and the project manager) feels reasonably certain that the task can be delivered within the

estimated time. Task times will be set with 80 – 90% chance of finishing within the set time. The logic is described in the graphic below (you read the entities from the bottom up). It looks like each task should have at least an 80-90% chance of finishing on time. If every task does finish on time, then

clearly the project will also

finish on time. Isn’t this what actually happens – every resource (and every project manager) includes enough time in every task so that he can be quite certain all (the vast majority) of tasks are actually completed within the estimated time.

I am quite sure that a large section of the project manager population and also task owners will be disturbed by the logic above. They must claim that estimates are much closer to the 50:50 chance of finishing on time. Their experience tells them that estimates are almost always insufficient – many projects have great difficulty finishing on time, on budget and with scope intact. Therefore a project manager’s task and project time estimates must include a lot of risk! Estimates are no better than 50:50 – at least that is the belief.

What is the truth? Are project managers people that manage risk appropriately – e.g. they create plans they believe they will meet with a high degree of certainty, or are they (mad) risk takers that produce task and project time estimates that are clearly unrealistic? Maybe the evidence should be checked! (I assume actions that can be taken to mitigate risk are implemented – this reduces risk, but cannot eliminate it – a part of risk management will always be to plan for ‘Murphy’).

There are plenty of reports that show that a high proportion of projects are delivered late. If this is true, then either project due dates are wildly

optimistic or, if task time estimates are realistic, then most tasks are finished within the time set for them and it takes just those few delayed tasks that cause a project to be late (or very late). This last sentence could be true if the majority of tasks finish on their due date (or very close to it)

and very rarely early. Check your task statistics. You should see that the vast majority complete on or near their due date (or they take the elapsed time given) and a few will be late or very late. If this is so, then the skewed curve above is not correct – there is something that is causing a huge spike on the due date with very few early finishes and a significant enough number of delays to cause many projects to be late.

If you can confirm that most tasks do finish on time - either because you or your company measure this performance or simply because that is your experience - please let us know. If most tasks are on time then the important question becomes,'why are so many projects late - or have great difficulty in finishing on time?'