Complex problems have no clear answers — but decision trees help. Learn how to define decision criteria and map outcomes.

There are four kinds of problems: Trivial problems, complicated problems, complex problems, and chaotic problems.

For trivial problems, there are clear goals and clear solutions. That’s not interesting.

Complicated problems have either clear goals but unclear solutions (e.g. technical problems), or clear solutions but unclear goals (e.g. political problems). That’s the majority of problems, and there is more than enough good practice for all sorts of complicated problems. So that’s not interesting, too.

Chaotic problems have both highly unclear goals and highly unclear solutions. Therefore, the solution finding will require novel practices, as such problems have never been seen before. That’s interesting, but not the purpose of this article.

So let’s stick with complex problems for the remainder of this article. Complex problems have both components of unclear goals and unclear solutions, but to a lesser extent than chaotic problems.

Let’s look into some examples.

Examples of Complex Problems

In entrepreneurial life, complex problems are all those problems to which you say: “There is no black-or-white answer.” Here are some examples from my experience as the Founder & CEO of Yonder, a B2B SaaS company.

Your product roadmap is a complex problem. What’s more important, building features to satisfy existing customers, or building features to attract additional customers? You will argue that both are important, but in any company, resources will be limited, so you will have to make a decision.

Your pricing policy is a complex problem, too. Should you decrease your prices to win over new business, or maintain the good margins of your existing business? Again, both approaches are important, but you won’t be able to have both.

Finally, attracting new investors is a complex problem, too. Usually, attracting the first investor in a round is pivotal for attracting further investors. So you will have to play your game wisely in attracting the right investor first. Much can go wrong, and a lot of if-then-else will be required to get your preferred new investor on board as the first one.

Decision Criteria

Because there are so many different complex problems out there, you cannot treat them all the same. That’s where decision criteria come in: Before you start working on a complex problem, define the decision criteria you will use to rate possible solutions for your complex problem.

For the product roadmap example above, the decision criteria might be customer satisfaction, RFP win rate, and technical complexity.

For the pricing example above, the decision criteria might be margins, liquidity reserves, and market share.

And for the investor example above, decision criteria might be strategic fit, ticket size, and terms.

Decision Trees

Decision criteria can be used to rate possible solutions for your complex problem. But they can also be used to create branches in your decision tree. Every decision criterion will be one node in your decision tree, and the different possibilities for each decision criterion create the branches in your decision tree.

Complex? Not at all. Let’s look at the roadmap example we used before in this article.

Developing the feature set requested by your existing customers first will increase customer satisfaction, but lower your RFP win rate. At the same time, you can also opt to develop the feature set requested by new customers first, therefore increasing your RFP win rate, but at the expense of your customer satisfaction. Irrespective of which path you follow, you can focus on either customer satisfaction or RFP win rate in the first step, but not both. You can use the criteria of technical complexity to choose the sequence of focusing on customer satisfaction or RFP win rate.

So as a result, your three decision criteria give you 2³=8 possible courses of action. That’s true for any three decision criteria, and if you go to four decision criteria, you will end up with 2⁴=16 possible courses of action, and so on.

Choose Your Desired Path

Once your decision tree is constructed, choose your desired path and then do everything to maximize the probability that the decisions branch along your desired path.

But never forget, no plan survives reality, and that’s the same for decision trees: Probabilities are never 100% and 0%, so always treat the undesired paths in your decision tree as contingency plans, fallback options, or BATNAs.