But the Problem Actually IS Complicated! Seven Reasons to Build a Model to Solve It

Have you ever had a conversation with a colleague that sounded something like this?

You: This is a complicated problem. There are a lot of variables to consider and many of them are interdependent.

Colleague: I get that, but you need to simplify it somehow or no one is going to understand it.

You: The problem isn’t simple, that’s why we haven’t solved it for the last year. We keep trying to treat it as a simple problem and when we try to  fix a part of the problem, we screw up something else in the process.

Colleague: I am just saying that by the time you get through explaining all this analysis, no one will be listening anymore.

Photo Credit: HowardLake via Compfight cc
Photo Credit: HowardLake via Compfight cc

You: So what do I do, just forget it and let things  blow up so someone will finally find the attention span needed to deal with it?

Even in this short exchange, you get the flavor of the frustration both people feel in these circumstances. One option is indeed to wait until things blow up and then the problem will get some attention. See my previous post on “leaving the donkey in the sun” for more on this option. But what are other options?

One approach to helping colleagues understand a complicated dynamic is to create an “interactive model.” For example, suppose that you have a complicated budget or financial investment problem with several interdependent variables like depreciation, inflation, incremental investment, and efficiency savings . Or maybe it’s a resource allocation problem with different skills sets, projects, deadlines, etc.

Create a spreadsheet with formulas that link the variables and demonstrate how making changes to one variable affects the others over time.

Here are seven advantages to building and using a model to solve a complex problem: 

  1. Instead of trying to explain all the complexities first, you can start by asking your colleagues how they expect one of the variables to change in the next year and let the model recalculate to show the impact on the other variables.
  2. You don’t have to advocate at the outset for a particular solution because now you have a method to test different approaches. This reduces the likelihood of immediate push back from colleagues who have strong biases towards particular solutions.
  3. You also have a method to test simplistic solutions and show how they have unintended side-effects because they don’t address the complexity of the problem.
  4. You can get your colleagues to help improve the quality of the underlying assumptions and data  in your model to make it more accurate. This may also make it more widely accepted.
  5. You can use the model to highlight trade-offs that may be required and align the choices on these trade-offs explicitly with the company culture and espoused values.
  6. You can also use the model to show how decisions that one manager makes impact activities managed by others. This will help ensure broader management support for dealing with the full complexity of the problem.
  7. Once you agree on a strategy to address the problem, you can keep testing the model against the actual results and keep adjusting the model and the strategy based on experience.

So when the problem actually IS complicated, build a model to help you communicate the complexity effectively and engage others in solving it.