I'm using a 4x4 matrix. I'm setting the scale, as a percentage of the dimension of the project under examination, in term of duration and cost. (I want to use "quantitative scale" for each risk cluster in order to facilitate the Risk Owner in the evaluation). It is the right approach or should I use standard percentage, if they exist? Do you have any tips for me when setting scale for both probability and impact? Thank you in advance. Saving Changes...
Sort By:
Luis BrancoCEO| Business Insight, Consultores de Gestão, LdªCarcavelos, Lisboa, Portugal
Dear Teresa
Interesting your question
Thanks for sharing
Are you comfortable using a 4x4 matrix for qualitative risk analysis?
How does your team react to this approach?
Are you satisfied with the results?
We usually use a 5x5 matrix, with the following parameters:
- Very low
- Low
- Moderate
- High
- Very High Saving Changes...
Teresa,
The approach comes down to stakeholder risk tolerance. There are probably dozens of standard percentages depending on who you ask, so there is no real standard. I typically use a 3x3 matrix for simplicity, or a 5x5 matrix for more differentiation, so even there you see there is no right answer.
When I joined one PMO as the new lead, they had no qualitative approach to risk evaluation, so my first task was to have discussions with our prime stakeholders to do exactly as you are asking. We had to define the levels. Even there, it becomes quickly clear that if you use the standards at the organization level, it will be quite different than at the corporate level. What is a big problem for me goes unnoticed by the CEO.
The conversations that generate your gradation and then how you evaluate individual risks is the most important part. Not everyone will agree on the precise answers, but most will have answers that are within one square on the grid. We may not agree on whether some problem is a 2 or a 3, but the absolute value is not as important as the knowledge gained in the evaluation. That is where I like the 5x5. There is less arguing about where it goes on the grid if we understand that close is good enough for practical use.
On the magnitude scale, it goes from extremely expensive to trivial depending on your organization's cost structure. On the probability scale, it goes from near certainty to near impossibility.
How accurate is your probability? Rule of thumb: You can't place anything more exact in location than the data you use to back that up. If your probabilities are "SWAG estimates", don't spend too much effort on exact percentages. Tolerance analysis is only as good as the first significant digit in the error. Saving Changes...
Unless you have good historical data to base probability and impact estimates on, you will be using a subjective evaluation approach leveraging expert skills.
As such, I'd keep thing easy with a 3x3 approach and use a conservative probability approach such as 10%, 25%, 50%.
Consider using a technique such as Delphi to get your stakeholders to simultaneously vote on their assessments of the probability and impact dimensions.
Kiron Saving Changes...
Sergio Luis ConteHelping to create solutions for everyone| Worldwide based OrganizationsBuenos Aires, Argentina
My recommendation is take close attention to @Keith comment above.On the other side, my recommendations if keeping the matrix as simple as possible, no more than 3 categories: high, medium, low. But the key thing is: you need a guide for people to decide what high/medium/low is and assign it to each risk. Here comes what @Keith stated above. In our case, we have a risk tollerance threashold for cost and other for time. Saving Changes...
First, keep it simple and start. after first draft, you will have an idea. Then work on that Idea. Also, Keith's comment may help. Saving Changes...
Ramachandran SwaminathanRegional Delivery Manager| Oracle Consulting IndiaBangalore, Karnataka, India
Teresa-
First of all let me admit that my suggestion is only based on my logic that I believe is good to apply. This is not based on experience or expertise
Coming to the problem, I would say that smaller value projects(In terms of Effort and $$) which needs strict monitoring you may need more finer matrix (4x4 or 5x5). The reason is that prob x impact will have greater bearing in the case of smaller value projects and can make or break it. So you need a finer measurement scale there. For bigger projects with large Efforts/$$ value we can use a smaller matrix for simplification.
So you may want to group the projects based on their size and apply the matrix accordingly
Please note that for measuring the weight of an ornament we need a finer weighing scale in terms of grams, but for measuring body weight of a person the scale can be in Kgs Saving Changes...
We have used a 5x5 (impact x likelihood) for some time now, and recently add two new variables; controllability and urgency. I think it is a little too much for most folks at the moment, but I like the idea of "urgency" that brings in the time dimension which is often over looked. Seeing a sea of heat maps on corporate reports, with no time dimension is very frustrating, and of course most are reporting green. I understand from your previous thread about qualitative risk management that you understand this concept, but I've seen risks mapped out on a timeline (x axis) as circles with the size of the circle representing impact and likelihood. The vertical axis would represent controllability. I personally like this approach, vs the traditional grid affair as it visually offers a much better method of displaying more of the relevant information in a concise and easy to understand format. Saving Changes...
Thank for sharing! I'm new at these themes and your support has been really helpful for me. As soon I will perfectionate the methodology will be a pleasure to share with you in order to continually improve. Saving Changes...
