How you can beat analysis paralysis

When you research a topic, a tech field for example, it is easy to get trapped in analysis paralysis. Why? Because there is a high degree of uncertainty. You probably keep asking yourself questions such as, "What's the best way to structure the field? How is the field even defined?," and so on. Then, you probably think, "Great, let's just get some more information". After all, having more information is supposed to reduce uncertainty, right?

So you keep searching, and collecting, and searching...

Then you check your watch, and you notice that you have spent half a day just searching and collecting. But you haven't decided or concluded what to do with your information, or what it could mean.

This is analysis paralysis.

Why having more information can hurt more than it helps

Before we talk about how you can quite effectively beat analysis paralysis, I would like to show you a somewhat counter-intuitive research finding.

In 1973, Paul Slovic wrote a paper, Behavioral problems of adhering to a decision policy (you can get the paper here). One of Slovic's findings was that having more items of information available leads to an increase in judgment confidence but not in judgment accuracy:

This is tricky. The implication of Slovic's finding is that it is easy to fool yourself into thinking that having more information will also improve your judgment accuracy.

By the way, these findings have made it into a classic book on intelligence analysis, Richards J. Heuer's Psychology of Intelligence Analysis (p. 54 in the book). You can download the whole book as a PDF here.

As you collect more and more information, your confidence in your judgment increases but your accuracy does not.

So, what to do?

To beat analysis paralysis, set an alarm clock to 20 minutes

Of the many things I myself have tried to beat analysis paralysis, this is by far the simplest and most effective tactic. Because it is so simple and as far as I can tell risk-free (what's there to lose? 20 minutes?), I really recommend you try it out.

Why 20 minutes?

First, it is 20 minutes because this is roughly how long your brain can focus on something. A good entry point into reading more about the science behind this claim is this Wikipedia article on "attention span".

There is even a time management method, called Pomodoro Technique (pomodoro = tomato), that is based on this (although it is 25, not 20, minutes in the classic Pomodoro Technique). Here is a great alarm clock for the Pomodoro Technique (image from Wikipedia):

How does setting an alarm clock help beat analysis paralysis?

Quite simply, if you only have 20 minutes, you have the finish line in sight. Of course, you can't research many topics comprehensively within 20 minutes. But this is not the goal.

Rather than trying to be comprehensive, the goal is to get a first useful approximation.

Let me break down what I mean by "first useful approximation":

"First"

It means just that. It is an initial step, more steps may follow. But even if you approach a topic in more steps, the key is that in between steps, you reflect, prioritize, refocus, and discuss with your team and colleagues. In other words, you can iterate and tread lightly, which is always a good idea when moving in uncharted territories (and by definition, with any tech field analysis you enter uncharted territories).

"Useful"

"Useful" means that your findings should somehow result in concrete action. Concrete action is the ultimate analysis paralysis breaker. A finding that triggers concrete action could be a company, for example, that makes something that you could buy instead of building it yourself, which enables you to innovate faster and stronger; or they might be competition. A research publication could be useful too, for instance, if it describes a method or an algorithm that you could implement in your own products. A market estimate could also be useful. Particularly if it helps you connect a technology to an application.

"Approximation"

This may be the hardest part to accept, particularly if one is a perfectionist. Yes, we know that we will most likely not get to 100% in 20 minutes, but even 20% of something is better than 100% of nothing.

This is about intelligence gathering and analysis, not about mathematical axioms.

(Quote adapted from a tweet by Christopher Ahlberg)

The nice thing about the alarm clock is that it really forces you to accept the "approximation" part, so you do not have to consciously fight the perfectionist instinct (which is usually what gets you into analysis paralysis). The alarm clock does this for you.

And what should I do when the 20 minutes are up?

Of course, this depends on your circumstances. But let’s assume that within 20 minutes you discovered (and perhaps even explored a bit) 2-3 findings (companies, research papers, market estimates, patents, blog articles, or clinical trials, for example).

Armed with this first useful approximation, you could now decide what to do with your findings, and prioritize your next courses of action.

Then, perhaps you decide to put in another 20 minute discovery sprint. In this case you will probably notice another nice aspect of the “alarm clock method”: it is a lot easier to schedule a 20 minute time slot than it is to schedule an open-end analysis session.

OK, now it’s time to go discover and explore. I hope you find this approach to tech discovery useful!

Check out the full article here.