We all have a bias that makes it hard to know if we’re successful

We are entering a new computer-generated era that will make us question everything we see.
We are entering a new computer-generated era that will make us question everything we see.
Image: Reuters/Stefan Wermuth
We may earn a commission from links on this page.

How’s it going?

If you’re running an early stage company, you probably have a strong emotional reaction to that question.

You might feel like things are going well right now. It’s a bouncy frenetic high that has you smiling into your morning coffee. Or, you might feel like things are going poorly. I don’t envy the green-purple, sick-to-your-stomach knot that’s currently twisted up inside of you. Regardless of which emotional reaction you’re experiencing at the moment, how much confidence do you place in it?

As a founder and current advisor to a number of startups, I’ve found over and over again that initial reactions to that question are often wildly out of sync with reality. We usually have a very strong feeling about how things are going, but those feelings are often disconnected from the underlying state of the business.

Instead, they’re mostly based on what’s easiest to recall.

In his book Thinking Fast and Slow, the noted behavioral economist Daniel Kahneman explains that when our brains encounter a difficult question like, “How well is my company doing?,” they use shortcuts to estimate the answer. These shortcuts are called heuristics.

One common heuristic our brains use is called the availability heuristic. Kahneman writes:

“The availability heuristic, like other heuristics of judgment, substitutes one question for another: you wish to estimate the size of a category or the frequency of an event, but you report an impression of the ease with which instances come to mind.”

In other words, when we use the availability heuristic, we decide a question not by systematic research and data analysis, but by judging how easy it is to recall specific examples that are relevant to it. For example, when we’re asked to evaluate a specific topic, say the danger of flying, we will try to recall examples of planes crashing. If it’s easy for us to do this, then we will judge flying to be dangerous. However, if a plane has crashed recently, we’re more likely to easily recall an example of one. And that means right after one occurs, we’re going to judge that plane crashes are more dangerous—even though the actual underlying rates of planes crashing have stayed the same.

So how does this apply to your startup?

Well, when you’re trying to figure out how well things are going, you’re likely to reach into your memory and pull out examples of good things and bad things that have happened recently. If it’s easier to recall bad things, you’re likely to feel that sick-to-your-stomach feeling and conclude that things are going poorly. If it’s easier to recall good things, you’re likely to feel good.

But in either case, that ease of recall is incredibly biased. It’s biased toward what happened recently. And it’s biased toward those recent events that are most visible and memorable.

That means that the feeling you have is incredibly skewed. It’s usually not a reliable indicator of anything.

This is a problem. Because in order to effectively manage your business, you have to effectively evaluate how well it’s doing. If you think things are going poorly when they’re actually going well, or vice versa, how are you going to make decisions?  You can’t.

Unfortunately, our brains work this way automatically. So we need to try to find ways to manage around this problem.

The most effective thing to do is to use data to replace our brain’s automatic reactions.

To do this, you just need to remember the Two Cardinal Sins of Data-Driven Early-Stage Entrepreneurship.

They are:

  1. Coming to conclusions without using data at all
  2. Coming to conclusions using too little data

When you’re not committing either of these sins, you’ll be in a good relationship with your business, and you’ll be much better prepared to adequately answer the question that we started with: “How are things going?” Let’s dive in.

Sin 1: Coming to conclusions without data

If you’re not using a formal, methodical system to make product decisions, measure your sales pipeline, track your VC meetings, evaluate your marketing efforts, etc., you’re doing something seriously wrong.

And I can’t tell you how many people at early stage startups fail at this. They keep all of their leads in their heads. Or they keep track of feature requests by memory alone.


You’ll end up building features for the customers that you just talked to instead of the features that were requested by the most customers or the most important ones. You’ll end up losing leads and not following up. You’ll conclude that you’ve talked to every investor when you really haven’t. Or that investors don’t like your pitch when really only the last two were negative on it.

When you build a system to track, you’ll be able to adequately understand the reality of the situation and the reasons behind your decisions: You’ll be able to say, “5 of our most important customers requested Feature X last month.” Rather than, “Feature Y seems important.”

Once you have a system, though, you need to understand how to drive it. That’s where Sin 2 comes in.

Sin 2: Coming to conclusions with too little data

At an early stage, you’re subject to ridiculous amounts of randomness. When a customer cancels, it’s likely to send you into an emotional tailspin. This isn’t working. This product sucks. We’re wasting our time.

But in reality, a single cancellation could be significant, or it could mean nothing at all. You need to wait and see.

So once you’ve started recording your data, set yourself a cadence to check in on it. Daily is way too frequently. You won’t really be able to draw any conclusions from it.

Weekly, or bi-weekly is much better. You’ll start to be able to see trends instead of just individual data-points, which means that when you draw conclusions about the data, your conclusions will be more likely to be accurate.

Once you have all of this set up, whenever you get that purple-green feeling that something isn’t right, you can actually do something about it.

Instead of relying on your memory to tell you whether venture capitalists are positive about your product, you can go back in count up all of the meetings you’ve had and say, “Hey, when I really look at it, people seem to like it.”

And that will help you see things more clearly, make better decisions, and answer confidently when someones asks you:

“How’s it going?”

Follow Dan Shipper on Twitter.