As product managers we make decisions all the time. What the priorities are, how to deal with a particular problem, how to motivate the team and so on. When we do, we like to think that we are objective and that our decisions are the outcome of rational thinking. Unfortunately, cognitive biases often meddle with our decision making process, leading to poor decisions and bad judgment calls.
What is a cognitive bias?
First, let us establish what a cognitive bias is. A cognitive bias is defined as “a systematic pattern of deviation from norm or rationality in judgment” .
The key element in this definition is the word systematic, as cognitive biases cannot be turned on or off, sometimes affecting our view of reality and sometimes not. Instead they constantly influence human judgment so that reasoning errors are reliably produced.
It is critical for product managers to be aware of cognitive biases and how they work, as they can steer product decision making away from data informed, rational argumentation. Some of the most common types of cognitive biases for product people are the following.
Confirmation bias is the tendency to seek answers that validate our beliefs and discard information that do not support our views.
In other words when a person hears only what they want to hear. It is especially problematic if that person is a product manager, as confirmation bias essentially limits their field of view and suppresses their objectiveness in the decision making process.
For example, when getting user feedback a product manager may pay more attention in reported issues that fall in line with their personal experience of using the product. Or a product manager with a technical background may have a preference to a technical solution that they are more familiar with. In both cases objectivity goes out of the window.
Overcoming confirmation bias takes being open to other views and accepting the uncomfort of eventually being wrong. It is a fine line to walk on, as product managers need to lead with conviction and be right a lot, so that they can earn the trust of others and make things happen. However, being too confident on one’s own views can reinforce the effects of confirmation bias, so keeping an open mind is strongly advised.
It also helps to be careful to avoid prescribing solutions to problems, even unintentionally. For example, a practical approach when talking to users would be to avoid asking: “How likely would you be to use this product/feature?” Framing the question this way nudges the user to think along the lines already presented. Instead ask users to describe in their own words the problem that they have. Thoroughly exploring the problem space before moving to the solution space will help to avoid prematurely prescribing solutions that validate our way of thinking.
The Sunk Cost Fallacy
The sunk cost fallacy occurs when people refuse to abandon a non viable endeavor that they have invested considerable effort in.
Consider the following situation: a product or project that month after month does not hit the mark, but instead of killing it more resources are poured into it in an effort to make it work. The reasoning behind these decisions is that since so much resources have already been spent, it has to work so that all this effort and money is not thrown away.
The key point is to be able to accept when it is actually time to quit something. It is not easy, as discarding a project is essentially admitting that a mistake was made in spending effort on it. But acknowledging a mistake is the first step in the process of fixing it.
There is not a clear-cut methodology to follow when deciding whether to push forward or whether to pull the plug on a non viable product. Such issues usually span beyond the decision making power of product managers themselves and are influenced by the overall business strategy. Organizational politics and personal egos also play a role. Getting this right can be a career booster (and likewise getting this wrong can be a career staller, at least in your current company).
A data driven approach, good knowledge of the market and the specific domain and alignment with the overall business strategy are advised for such cases.
Survivorship bias is the tendency to draw conclusions based only on what has worked in the past (ie. has “survived” a process) and ignoring what has not.
Under the effects of survivorship bias one aims to harvest the winning elements of past wins, subconsciously (or not) trying to reuse the same formula for success. However, this line of thinking ignores the fact that a past success may be due to luck or that it was influenced by different environmental conditions. The different conditions depend on the specific context and could range from market related like competition maturity or pricing, user related like different user segments or cohorts, organizational related like different team dynamics and so on.
Thinking this way also pushes away the opportunity to learn from past failures, opening the door for the same mistakes to happen again.
For example a product manager under the influence of survivorship bias may concentrate only on current users of their product (that have “survived” the onboarding process) and ignore churned users. Doing so they miss the opportunity to learn why the rest users chose not to continue using the product and the possibility to get insights on how to reduce the churn rate. Or they may support a specific solution to a problem that has worked for them before, disregarding the fact that the context of the problem or the conditions may be different.
The Curse of Knowledge
The curse of knowledge occurs when an expert on a topic assumes that other non-experts have the background to understand.
This is a phenomenon very relevant to senior product managers who live and breathe how their product works and often assume that users have a comparable level of knowledge about it. A product manager affected by this cognitive bias could wonder when facing users that have trouble using it: Why don’t they get it?, forgetting that all the product exposure that users may have is a superficial view of the main screen and (maybe) a glimpse of the manual.
Another aspect of this effect is that product managers often tend to favor feedback from their advanced users. Advanced users can understand the intricacies of how the product works and can engage in deeper conversations about it. Product managers have an easier time communicating with someone who can “speak” the same product language, without the need to explain the basics. The result is that more effort is directed towards meeting the needs of advanced users, increasing complexity and making the product more difficult to use for new users.
To overcome the curse of knowledge product people need to practice empathy when talking to users. They need to consider the level of product knowledge that users have and adjust their communication for their specific level of understanding. It is important to understand that users have different needs depending on where they are in their product journey. Disregarding the needs of a specific segment, for example new users, will prove detrimental for their experience and will turn them away from the product.
Cognitive biases have the potential to affect decision making in ways that are difficult to discern. When they do, it is often hard for the affected individuals to accept that their way of thinking is systematically affected. Product managers as decision facilitators are especially prone to be affected.
Fortunately, there are actions that can be taken to minimize the detrimental effects of biases. The first is developing awareness. Being aware of what cognitive biases are and in what way they can affect judgement is key in the process of overcoming them. The second is vigilance. Actively monitoring against common biases will help to detect biased decision making early, allowing to course correct before any major damage is done and minimize the impact of any cognitive bias at play.
: Haselton, M.G., Nettle, D. and Andrews, P.W., 2015. The Evolution of Cognitive Bias. The Handbook of Evolutionary Psychology, pp.724-746.