Why Complex Decisions Inevitably Take Weeks

convergenceI often find that, when it comes to make certain types of decision in an organisation, this just seems to take weeks. And if you’re unlucky, this can roll in to months. Why? What is it, a lack of decisiveness? An unwillingness to commit to anything? Lack of identification of a “Decision maker”? Just weakness!? I suggest that it’s none of these things, but in fact inevitable, when it comes to a certain classification of problems. More specifically, with complex problems (definition below), there’s actually very little you can do, but accept the time it takes, and stop worrying.

Here’s an example, of the sort of discussion that’s happening in offices up and down the country. NB: this is purely for illustrative purposes, and not related to any real events!:

  • Colleague 1 – “We’ve got a problem hiring account managers. Let’s have a meeting to figure out what we’re going to do about it”.
  • Three colleagues sit and discuss the problem and decide “We’re going to increase the salary on the website, to attract more people”
  • Colleague 1 tells the HR team. At which point the Head of HR points out “What do you think our current account managers are going to think when they see that salary on the website? We can’t do this – let’s have a meeting”
  • Colleague 1 and Head of HR have meeting and decide “We’ll increase salary for the role, but we’ll put a very wide range, to keep current employees happy”.
  • New salary range goes on job ad, on site
  • Applications start coming in – from lots of junior people who assume it’s a junior role, from the low starting salary
  • At the same time, account manager colleagues start asking “Why aren’t I getting that top range salary? I did an amazing job last year!”
  • Various colleagues and HR people sit and have another discussion about the salary range, and decide to tighten it up again, to stop junior people applying and to keep employees happy
  • Update page goes live
  • Colleague 1 chips in  – “Hang on, we’re back where we started! We still haven’t solved my problem of getting an account manager – what’s going on?” 

This isn’t supposed to be an example of the difficulties of hiring. It’s an example of a long decision making process going round and round in circles, leading I suspect, to frustration amongst all of those involved.

But why has this happened? Why is it so difficult? Fundamentally, it’s because the problem of “How will we get more account managers in the building?” is a complex issue – and therefore has different characteristics to simpler problems.

The Harvard Business Review article A Leader’s Framework for Decision Making outlines a framework for classifying problems, crudely summarised as follows:

  1. Simple Problems – simple cause-and-effect problems, often with obvious answers. And often something that happens all the time. The example I like best is “What to do if someone comes in to complain about your hotel?”. Is this a difficult problem, requiring a committee of your greatest minds to figure it out? No! Just give the guy some kind words, may be a discount code and keep him happy. It’s a simple problem, and generally “Best Practice” should be applied.
  2. Complicated Problems – again, there is cause-and-effect here, but more expertise is needed to fix the problem. The example in the text is a problem with your car engine. There is a cause-and-effect going on here, but what is it? Can anyone fix it? No, you need expertise and it might take hours to diagnose the problem. Software Engineering can also, often, fit in to this category – okay, you want the website to do X and Y, and it’s certainly possible to do this, but it might take years of experience and expertise to make that happen.
  3. Complex Problems – the key difference between complicated and complex problems is that, for the latter, the domain is simply too complex and unpredictable to undertake a priori analysis and come up with the answer in one shot. There are feedback loops – making one decision impacts other parts of the problem space, which feed back in to the original decision and change the context. In the example above, increasing the salary range on the site causes disgruntlement amongst employees, leading to a revision of the original decision and so on.
  4. Chaos. No-one knows what the hell is going on, and trying to sit back and rationally analyse cause-and-effect is pointless. Your job is to “Stop the bleeding” and manage the problem actively.

So, our example is a “Complex” problem – and I’d argue that a large number of management problems fall in to this category. Why? Because you’re dealing with people, their aspirations, irrationalities, emotions and so on, and these are unpredictable at best. Or even without the need to take peoples’ responses in to account, many problems still fall in to the domain of the complex. We had a recent process to try and figure out a new pricing structure for our products and suites of products, and this took weeks to figure out. Why? Because all of our products, functionality and bundles are weaved together in a web of interdependency – decreasing the price of product X makes bundle A a more difficult up-sell, but that makes the prices of other products in the bundle up for change, which then affects bundles B and C and so on, and so on. This took weeks to resolve as we went round and round trying to reach a stable equilibrium which made sense.

So, making decisions like this are always multi-step. And worse, almost always involve lots of different people (particularly in an org where consensus building is important – a different topic, to come!). Many of whom won’t be available at the same time (people are busy), who often need different levels of information, already have different levels of knowledge of the domain space and so on.

A very large number of problems fall in to this type – where meeting 1 leads to a decision which then affects others, leading to meeting 2, which then affects the decisions in meeting 1, leading to meeting 3 and so on and so on. What you’re trying to do is reach a stable equilibrium point – where, all things considered, most people are kind of happy with the outcome and all of the issues have been considered (many of which won’t have come out till meeting 2 or 3). NB: This is also why these decisions can often feel like compromises – because they are, and the better for it! There is no simple, obvious-to-all answer, a silver bullet which we should have realised on day 1.

So next time you’re getting frustrated with a decision making process, asking your colleagues “Why is this taking so long? Just make a decision already?” take in to account that it’s likely to be a complex problem which can’t be forced.

What you can do of course, is create an environment where the process happens as smoothly as possible – are the right people involved at different points? Are there people making the problem worse? Do you have people who can work through complex problems, with poor quality data, and intangible concepts? Can you yourself facilitate the process (rather than jumping in with the magic answer)? On this last point, I like the comment made by Kishgore Sengupta*:

Instead of saying ‘Don’t just stand there, do something’ your new behaviour should be ‘Don’t just do something, stand there'”


* I’d like to thank Cambridge Judge Business School for introducing me to these concepts, particularly for the course given by Kishgore Sengupta on complexity – fascinating stuff!

Read More