Decision Matrix Analysis

June 1, 2018

When there are too many variables and the right choice is not clear.

Decision Matrix Analysis; also known as Pugh Matrix Analysis, Grid Analysis, and Multi-Attribute Utility Theory is a very simple and useful tool when faced with many choices, each having several pros and cons.

Its value lies in two main areas; Methodical Organization of choices and Weighting of the Criteria that are most important to the deciding party.

Methodical Organization

What is the first thing most people do when you are overwhelmed by things? You make a list. Shopping lists, packing lists for vacation, a list of tasks you hope to achieve for the day. But lists can also be useful when making a choice. It really doesn’t matter if it’s choosing the right vendor for your business, a college for your child, or a car for yourself… the first and best thing to do is make a list of realistic finalists.

A Decision Matrix does just that. It lists the choices down the left side of a spreadsheet or piece of paper with a grid on it. Since we are business consultants and this is geared towards business owners, we will use business vendors for our example… but please imagine it could be colleges, cars, or even mates!

On the top we add the criteria or factors by which we will measure or judge our list of vendors.

Decision Matrix
Criteria Cost Reputation Service Speed
Big Boys

If you had to choose between 20 vendors and there was a dozen criteria, you can see where the Methodical Organization would be very helpful just to keep track of it all and see it all on one piece of paper. But if you only had 5 vendors based on 4 criteria like the above, you could probably do it in your head and are wondering… why would I waste time on this silly grid?

Weighting the Criteria

Giving the right amount of importance (weighting) to each criteria is very difficult for most people to do simply because of our subjective human nature. In our hearts we LIKE one vendor more than the other. Maybe we have used them before, or have a friend there, or just plain like their name better. It is very easy to ignore the whole matrix and just go with your gut.

Assigning weights forces you to compartmentalize each choice like taking a multiple choice test… you focus on one question at a time and at the end you see your score. That “one at a time” focus removes all subjective biases and forces you to objectively answer each question. The end result is the objective truth!

How To Do It  

Rate each vendor in each criterion from 0 to 5 with 5 being best (If you had 6 choices it would be 0 to 6, and so on). For instance, if you know Zepher is the most expensive put a 0 in their cost box. If ACE is the least expensive put a 5 in their cost box. You don’t have to allocate 0 through 5, if a vendor is bad at all criteria they can be 0 across the board.

Decision Matrix
Criteria Cost Reputation Service Speed
ACME 3 4 5 5
ACE 5 3 2 4
Big Boys 2 2 2 2
Macrosoft 0 0 0 0
Zepher 0 5 5 5

After rating each vendor for each criterion, we add a weight to each criterion from 0 to 5. That is… which criteria are not important (0) and which are very important (5). Like the rating, you don’t have to allocate 0 to only one criteria and 1 to another. There can be two criteria that are equally very important.

Next, multiply the weight in each category by the rating you gave each vendor and replace the rating with the weighted total.

Lastly, add up the criteria score for each vendor.    

Decision Matrix
Criteria Cost Reputation Service Speed TOTAL
Weight 5 2 4 5
ACME 3  15 4  8 5  20 5  25 68
ACE 5  25 3  6 2  8 4  20 59
Big Boys 2  10 2  4 2  4 2  10 28
Macrosoft 0 0 0 0 0
Zepher 0 5  10 5  20 5  25 55

You can see how ACME won but had you cared a little more about reputation and little less about cost… Zepher might have won.  

The Missing Piece   

You may be wondering “Yes I get all the math, but how do I determine my criteria and worse yet… how do I find out how each vendor REALLY is for each criteria?” The answer to that is what management consultants do all day, every day… research!