I have always been interested in decision making and how people often are not logical in not only their preferences but even how they remember and look at facts. The most useful book I ever read was Thinking, Fast and Slow by Daniel Kahneman, (highly recommend it if you haven’t read it yet) and one of my favorite academics is behavioral economist Dan Ariely. Not only does Kahneman and Ariely’s research help you understand consumer behavior, it helps you understand your own decision making and, most importantly, mistakes most of us make.
A recent guest blog post on the Amplitude Blog, 5 Cognitive Biases Ruining Your Growth, does a great job of describing five biases that can greatly impact your business. While I will try to avoid just repeating the blog post, below are the five biases and some ways they may be impacting you:
- Confirmation bias. Confirmation bias is when you interpret or recall information in a way that confirms your preexisting beliefs or hypotheses, while giving disproportionately less consideration to alternative possibilities.This bias occurs regularly in the game space, especially with free to play games.
A product manager may have driven a new feature, maybe a new price point on the pay wall. Rather than running an AB test (maybe insufficient traffic or other changes going on), they then review the feature pre and post launch. Game revenue per user increased 10 percent so they create a Powerpoint and email the CEO that there new feature had a 10 percent impact. Then the company adds this feature to all its games. The reality is that at the same time the feature was released the marketing team stopped a television campaign that was attracting poorly monetizing players. The latter is actually what caused the change in revenue. As someone who has known a lot of product managers, I can confirm this bias in the real world.
- The narrative fallacy. People try to comprehend information in stories, rather than looking at just the facts they create a story that links them together even if there is not really a link. If you watch business news, when the stock market goes up 5 points, the narrative may be the market has rebounded from its Brexit blues. If the market goes down 5 points, the same story would be the market is still suffering from Brexit. The reality is that 5 points is statistically insignificant (the market is an aggregate of multiple stocks) so neither narrative is more likely in either scenario. The key issue here is that we attribute causation where there is none.An example in the game world.
Two branded games are in the top 5 of new releases. All of the analysis is that branded games are now what customers are looking for. The realities is that the two games, totally unrelated, had strong mechanics and were just that lucky 10% of games that succeed. Allowing the Narrative Fallacy to win, however, you then put your resources to branded games, which are no more popular than before the launch of the two successful titles.
- Dunning-Kruger Effect. Before the Amplitude post, I had not heard of this bias, at least with this name, but once you read about it I am sure you will know cases of it. The Dunning-Kruger Effect is when incompetent or somewhat unskilled people think they are more skilled than they are. As the article quotes, “incompetent people do not recognize—scratch that, cannot recognize—just how incompetent they are.”
Again, for the example from the game industry. Let’s say you want to port your game to a new VR platform. You go to your development team and they say it won’t be a problem. You sign up for the project, give them the specs, six months later they still cannot get the game to run on the VR platform as they have no idea how to develop VR (this is a nicer example than some others I can remember).
- Backfire effect. The backfire effect is after analyzing something that you or your company are doing, if the results are negative and the action was bad, you or your colleagues refuse to accept the results. As they write in the blog post, “the exact definition of the Backfire Effect [is]: ‘When people react to disconfirming evidence by strengthening their beliefs.’”
As an example, you decide to analyze how your company has been calculating LTV. You look back at the analysis done the last two years and see how actual LTV tracked with projections at that time. You discover that you underestimated actual spend by 50 percent. Should be great news, will allow you to ramp up dramatically your user acquisition. Instead, when you present this data to your analytics team, they refuse to accept it, saying your analysis is flawed because you are not looking at the right cohorts.
- Bandwagon effect. The bandwagon effect is what you would assume, the tendency to do things because many other people are doing it. People will rally around a cause, an idea, a candidate, a variation, or a strategy simply because it is popular.
Given that I want to keep this blog post under 500 GB, I will not list all the examples of the bandwagon effect I have seen in the game industry. Product strategy, however, is the most obvious culprit. When the free to play game industry started to evolve to mobile, everyone started porting its Facebook games over to mobile. Since Zynga and the other big companies were doing it, all of the smaller companies as well as newly funded ones also tried to bring the same core mechanics from Facebook over to mobile. Mechanics that worked on Facebook, however, did not work on mobile but companies continued doing it because everyone else was. Rather than identify the market need and a potential blue ocean, companies just joined the bandwagon.
Avoid these biases
The key to making the right decisions is not to assume you do not have biases, but always to be diligent in reviewing your decisions and making sure you are thinking rationally. All of these biases can lead to personal or company failure, so the inability to identify them can have extreme consequences.
- Understanding our biases allows us to not only understand our customers but make better decisions.
- A core bias you see in the game industry is confirmation bias, where someone looks at data to prove their hypothesis (or brilliance), even if the data does not really support it.
- Another critical bias is the narrative fallacy, where we create a story to explain an event even if the story is not the cause of the event.