In the previous article I discussed some strategies for solving problems. I discussed why it is important to understand the *why* of the problem, and the *what* you need to achieve.

This article looks at the *how* of solving a problem. It leads into why having several plans to resolve the problem is a good thing.

Given you know the why and what, and have retired assumptions and constraints, or even added some, how do you go about pursuing a solution?

To answer this we turn to Edward de Bono, who proposed some very interesting approaches, considered quite radical at the time. He is the protagonist of *lateral thinking*, which is now regarded as the seminal way to approach problem solving.

First, you must understand that the general approach sees you latch onto the first solution that comes to mind, and pursue it, looking for reasons to accept or reject it. Under de Bono’s method, you do something very different. Having thought of an approach you park it immediately, and start to contemplate other ways to solve the problem. This is described as a breadth vs depth approach, where the breadth of solutions is much more important than the depth of any one solution – hence the name lateral thinking.

What you should get are several possible approaches. What is important is that you have broadened your thinking on what might be an effective way to solve this problem.

This process of proposing solutions is best done in a group. You should include those people whose opinions you value, and more importantly, the team members who may have to work on the solution. Whatever group you choose, unless you have both objectively independent views and subjectively enrolled views you are limiting your options. You can, of course, do this as an individual, but management studies show that a group approach produces far better results.

Once you have a set of possible solutions it time to look a little deeper at them. What you are looking to do is not refine the approach, but to determine whether it would lead to a desirable solution. You may reject approaches because of time or cost, but to discard based on politics, conventional thinking, or other subjective matters is immensely limiting, and will lead to a mediocre solution.

At this stage you should have several possible approaches, any of which could produce the outcome you need. If you have less than three, I suggest you haven’t explored the matter well enough. If you have many more than ten, then I suggest that some of your approaches may be similar, just differing in detail. These should be re-examined to see if they are not the same approach, just couched differently. If they are just simple variations, then group them together.

What you do now is rank these solutions. Just asking your group to vote on a favourite will not produce the results you want. Favourites voting will likely just have your stakeholders voting for the idea they proposed, and you will arrive at a conclusion that all possible solutions are equally preferred.

You should use a technique called the *analytic hierarchy process*. It was developed by Thomas L. Saaty in the 1970s and has been extensively used since then.

What this technique proposes is a pair-wise comparison, and asks people to nominate their preference out of each pair. Besides ranking each option relative to the other options, this approach will minimise the effect of each individual’s bias towards their own solution.

To use this requires a little mathematics. There are n*(n-1)/2 pairs present in your solution set, where n is the number of solutions. Given say, six solutions, there are 15 pairs to be evaluated. If you have five participants in your planning session, then that means you have 75 observations to accumulate and analyse. A spreadsheet is your friend here – and you will not be surprised to learn that there are several templates for this analysis. At the bottom of this article is a link to one I have used often. It is a Microsoft Excel template.

After you crunch the numbers, you will have your solutions ranked, based on the team’s preferences. Clearly, the top of the list is your Plan A, and you should leap straight into it, right?

Well, actually you shouldn’t. Before doing so, you need to understand confidence limits and their relationship to Plan B.

If Plan B was any good it would be Plan A.

IT project axiom

I only agree with the above statement if you haven’t properly premeditated a Plan B. Generations of operations researchers have come to the conclusion that if you only consider Plan B once Plan A has failed, then you will almost certainly fail with Plan B too.

The right time to consider Plan B is whilst you are formulating Plan A. This seems counter-intuitive because you probably think you should throw all your resources at Plan A to make it succeed. This is only true if you are completely certain that Plan A will succeed.

If there is the slightest doubt, then Plan B should be exercised at the same time as Plan A.

Clearly this means that you need to deploy resources, be they people, time or money, away from Plan A to Plan B. The question is how much?

The answer lies with your confidence in Plan A providing a solution. Those generations of operations researchers I mentioned have produced some guidelines for determining the optimal division of resources given your confidence limit. If the mathematics behind this are of interest, see the further reading section below.

The answer lies somewhere between 37% and 9%. If your confidence is at 50%, then you should devote 37% of your resources to Plan B. If you confidence limit is around 99%, then you only need to deploy 9% of your resources to Plan B. Above 99% probably means you don’t need a Plan B, but you had better be sure. If your confidence limit is below 50% then it suggests that this probably shouldn’t be Plan A, because you are saying that is has a greater chance of failing than succeeding.

There are times when it might make sense to take on such a risky approach, but do so in the clear knowledge that it is risky and likely to fail. But, you have a good, developed Plan B, right?

If those numbers, 9% and 37% seem familiar, then you are right. Euler’s number, or *e*, is in play again. 37% is 1/e, and 9% is 1/(4*e).

For most purposes, you can assume that the Plan B allocation is a straight line between the 50% and 99% marks. It isn’t, but given the coarse estimation of confidence in your plan, then it will be close enough.

What happens when Plan A fails and you switch to Plan B? You do exactly what I have specified above, but bring Plan C into consideration, and deploy your remaining resources accordingly.

But, I do hope that your boss, who has you as Plan A, has a Plan B themselves.

### resources and further reading

Further reading on Edward de Bono and his techniques:

www.edwdebono.com/

For more on Saaty’s decision-making process, the following paper summarises his concepts.

Decision making with the analytic hierarchy process.

Excel template for analytic hierarchy process calculations:

Business Performance Management Singapore