We all make a lot of decisions, big and small. Most of the time we don’t need to over-think the process, right? I mean, if the printer is out of paper, just get more paper. No need to construct a T-chart with pros and cons. Sometimes, though, the effort it takes to build out a structured decision process totally pays off.
Here’s a method I’ve used a bunch of times, in a bunch of situations. Though I’ve used it in many contexts, I find it to be particularly helpful during a hiring process, especially with a group. It takes time to set up, and it takes even more time when you’re working toward consensus with a group (a whole different blog post), but if the decision is important, it’s time well spent.
Brainstorm criteria: whatever your decision is about, brainstorm a list of no more than a dozen criteria by which to judge your alternatives. So, if you’re looking at painting your living room, you might have a list like this: low VOC, high coverage, low carbon footprint, low cost, easy to source.
Pair rank your criteria. Take your list of criteria, and compare each one to all others. Decide, one at a time, which is more important to you. So, between low VOC, or high coverage…which is more important? Now compare low VOC against low carbon footprint: if you could only choose one, which would it be? As you run through the list, put a hash mark next to the criteria that “wins” each comparison. By the end of the exercise, you’ll have a clear view of which criteria is most important to you.
Weight your criteria. If you initially brainstormed a list of 12 criteria, you’ll see that some criteria didn’t get any “votes.” You can eliminate these from consideration. You might only want to consider criteria which gathered 3 or more hashmarks, or “votes.” So of the original 12 brainstormed criteria, maybe you’ve got 5 remaining criteria with at least 3 hashmarks. Add up all the hashmarks for all remaining criteria. Let’s say there are 18 hashmarks. If “low VOC” has 5 of 18 hashmarks, then it’s weight will be 5/18, or 27% (5 divided by 18, multiplied by 100). Do this calculation for each remaining criteria, rounding as needed so that your final percentages add up to 100%.
Evaluate alternatives against criteria. Now this gets a little technical. A spreadsheet is probably best for this step, since you’ll be arranging stuff vertically and horizontally. Since you’re usually more limited along the horizontal view, use whatever you’ve got the least amount of, criteria or alternatives, along the top row. Whatever there’s more of should be arranged vertically along the left side. Now you want to compare each alternative against each criteria. So, looking at Paint A and Paint B against the “low VOC” criteria, which would you choose: Paint A or Paint B? Now compare both against the next criteria, and so on. When done, move to Paint B against Paint C for low VOC: which would you choose? As before, as you make your choices, place a hashmark in the intersection of alternative and criteria to indicate your “vote.”
Calculate results. Now that you’ve brainstormed your criteria, pair ranked them, then pair ranked your alternatives against your criteria, you should start to see a decision emerging. At this stage you’ll need to multiply each alternative’s total hashmark count by the weight of each criteria. When finished, add up all the numbers accumulated for each alternative. You should see a clear “winner.”
Gut check. Once you’ve done all the math and you see where it leads you, how do you feel about it? As much work as this entails, it is an imprecise method. There are plenty of opportunities to make pair ranked selections based on emotion, rather than fact, and that can lead the entire process astray. Take a look at your results. Does it feel about right to you? If you or your team have any unresolvable reservations about the outcome, scratch it and start over. It’s a pain, but if the outcome is important, it’s worth the investment of time.
I know this looks like a lot of stuff. And for run of the mill decisions, it’s too much overhead. But if a decision matters, and you want to be able to justify your process down the line, this kind of process is pure gold.