Oxford University Press's
Academic Insights for the Thinking World

How to use maps to solve complex problems

Imagine that you’ve just been appointed the head of operations for a five-star hotel in Manhattan. Your boss calls you in her office on your first day and says: “Our biggest problem is how slow elevators are. Everyone complains about it, and we can’t have that. Speed them up.”

How would you do it?

Most people intuitively focus on speeding up the elevators: installing stronger motors, modifying the dispatch algorithm, getting the doors to shut faster, or even installing additional elevators. But there is also a whole set of options that consists of giving the impression that the elevators are faster, for instance by distracting the users with windows or, on the cheaper side, with mirrors, TVs, or newspapers.

Facing a complex challenge, it’s tempting to pursue whichever option we first identify and start implementing it quickly. But relying on this kind of intuitive thinking, particularly in unpredictable environmentsmight result in poor outcomes —even when we have a high degree of confidence.

Instead, we should first identify the range of solutions that are available to us, which is challenging, because we don’t know what we don’t know. It’s like being an explorer a few hundred years ago who would land on an uncharted piece of land having to decide how to explore it. Well, there is a time-tested approach: map it.

Facing that complex challenge, your first task should be to frame your problem in the form of an overarching question that, once answered, will lay out your proposed plan of action. A good way to frame your problem is to summarize that key question and its context in a situation- complication-question sequenceOnce you’ve framed your problem, you should then think of alternative ways—or options—to solve it before deciding which one is on balance the best.

A question map helps you explore that universe of potential options. Question maps are similar to other knowledge cartography tools, such as mind maps, decision trees, and the issue trees that management consultants use. But question maps are also different from these tools in that they obey their own rules, four rules that actively drive structure into your problem-solving process.

  1. 1. Good question maps answer a single type of question
    The first rule of good question maps is that they answer a single type of question. Facing a complex problem, we face two types of questions: why and how. A why question is problem centric; it helps you to diagnose your problem, uncovering its root causes and, ultimately, enabling you to summarize your problem in a more insightful key question. In contrast, a how question is solution centric: It enables you to consider all the possible solutions to your problem.One type of question isn’t consistently better than the other. Both are useful and, in general, you want to ask why before you ask how.
  2.  Good question maps go from the question to potential options
    The second rule of good question maps is just as simple: Start your map at your key question and explore all possible answers—the so-called solution space. This requires you to think in a divergent pattern. Vertically, use your map to uncover the various facets of your question, with each branch addressing a different dimension. Horizontally, use your map to go into more details eventually identifying concrete, tangible answers.So, if your question is how you should make your clients happy with the speed of your elevators, you might start a map with two ideas: “by speeding up the elevators” and “by creating the impression that the elevators are faster.” You can then continue exploring these ideas into further and further details.

    You should continue that divergent thinking pattern as long as doing so brings value. This will typically result in a map with dozens of ideas over at least five layers. For a why question, this progression in further depth is equivalent to the five-whys methodology—but don’t limit yourself to just five, if it makes sense to drill deeper, do so. Since it would be intractable to treat each one of these ideas as a separate option, you then converge by summarizing your entire map in a set of two to ten formal options.

  3.  Good question maps have a mutually exclusive and collectively exhaustive structure
    Next, consider all the possible answers to your question exactly once. That is, give your map branches that are mutually exclusive and collectively exhaustive (or MECE). Giving your map this structure is a critical element of the process, and in most cases, it isn’t easy to do, as it requires you to think both creatively and critically. So, expect to spend some effort here—and by “expect to spend some effort,” I mean that if you don’t feel some frustration, you’re not doing it well!
  4. Good question maps are insightful
    The final rule of good maps is that they should have an insightful structure. That is, the structure should be both logically valid and useful. Concretely, this means finding a way to structure the map that helps bring some new, useful light into your question. This is especially important for the first cut, the top-most node of the map.

Striving to be insightful also means that you need to keep developing your map until you identify concrete answers. Don’t stay at a philosophical level (e.g., by just writing “by creating the impression that the elevators are faster”) but identify concrete ways to answer your question (“by installing mirrors in the elevators”).

 

Question maps only have four rules, and those are simple enough to understand. Following them when solving a complex problem, however, can be, let’s say, challenging. But realize that the complexity isn’t in the map; the complexity is in the problem, all that the map does is expose that complexity so that you can better address it. So, embrace the challenge, and happy mapping!

Feature image by Daan Stevens via Unsplash

Recent Comments

There are currently no comments.

Leave a Comment

Your email address will not be published. Required fields are marked *