I’ve had 3 or 4 different CEOs show me this method for group prioritization lately, and I’ve since used it with two clients. I couldn’t find it on the web, so I might have the name wrong. If you know of another name for this technique, please drop a note in the comments.
Here’s the setup:
You are trying to prioritize something: Annual or quarterly priorities for your company, motivations behind taking a certain action, core values, anything. Use any method you like to winnow down your options to 5 (some people swear by 5, but I’ve also used it successfully with 6, maybe it works with any number?). Scroll down to the bottom to see the finished product, if you’re in a hurry and don’t want to see the sausage being made.
Here’s an example using our choice about kindergarten options for our two boys. We have many things we want from an opportunity. What’s most important? We winnowed down to these top 5:
- multi-age
- enough structure
- enough flexibility
- enough time for the boys to be away from each other
- low likelihood of differences to be labeled as behavior problems
Next step is to write those 5 down on paper or whiteboard, in a circle:
Then, pick one at random (time apart, in this example) and for each of the other 4 (one at a time), ask which is more important, or which is cause and which is effect. Draw an arrow from cause to effect, or from more important to less important. After the first round it looks something like this (time apart is more important than multi age, enough structure is more important than time apart, etc:
Now pick a 2nd item (multi age) and do pairwise comparison again (enough structure more important than multi age, etc). Note that you only have 3 comparisons this time because you’ve already done time apart vs multi age.
Keep going until there is an arrow between every pair of items:
Count the number of arrows pointing in. While it is possible to have ties, if you don’t you now have a simple prioritized list. Zero is most important and 4 is least important. If you have ties, then you have ties in your prioritized list. In our case, low likelihood of differences being labled behavior problems is most important and multi age is least important (of these top 5).
I recommend giving this a try on your own before you try doing it with your team. Let me know how it works!