3 cognitive biases Product Managers should be mindful of

And how to avoid them

At Booking.com product managers are highly data-driven. But even with a wealth of data at our fingertips, we’re just human beings prone to bias. On a daily basis we juggle many projects and liaise with several stakeholders. With so many competing priorities, it’s important to be aware of the ways cognitive biases can affect our decision making and ultimately the product.

Below are 3 cognitive biases product managers should be mindful of.

Sunk cost fallacy

We tend to think we’re rational human beings that only invest energy in features that bring value. Wrong! The truth is that our decisions are sometimes emotionally driven by our prior investments.

Imagine this: you prioritised a story, your team poured weeks of design and development into it, but it hasn’t delivered much benefit for the user or impact for the business. What do you do? The obvious answer is to abandon the story and focus efforts on more fruitful initiatives. In reality, people have a tendency to continue pouring resources into things in which they already invested energy — no matter how irrational it may seem. The more time, money or energy we invest in a feature, the harder it becomes to give up on it. This is called the sunk cost fallacy.

How to avoid this?

1. Size opportunities. Do a cost-benefit analysis to have a clear understanding of whether the benefits for the customer outweigh the time spent on development and design.
2. Take small steps. At Booking.com, we prefer launching MVPs rather than spending weeks building a perfect solution that may or may not work.
3. Always question your own decisions! 🤔

At Booking.com, product ranking on the home page is powered by machine learning to deliver the best ranking experience for each user.

Order bias

Order bias occurs when the order of features on a page influences the way the user interacts with everything else. When product managers launch a new feature, they usually want it to be seen and used by as many users possible. This may lead to product teams adding their feature at the top of the page, above all others. Although this is great in gaining exposure for the new feature, this may not always be what the user is looking for.

How to avoid this?

1. Put yourself in the users’ shoes. Understand what information the user needs first. It may not be your feature. 😉
2. Embrace machine learning. At Booking.com feature ranking is powered by machine learning to deliver the best ranking experience for each user.
3. Ask other product managers to be mindful too. ☮️

Confirmation bias

Sometimes people are so convinced of their own beliefs, that they subconsciously cherry-pick insights that align with those pre-existing beliefs, while ignoring others. This is called confirmation bias. In A/B testing this bias can manifest itself in product managers conveniently picking metrics that enable them to prove their hypothesis, rather than considering the bigger picture.

Imagine this: your team built a feature which delivers loads of bookings, but increases customer service tickets by a lot more. Would you consider this feature successful? If you would only take into account bookings, the answer is ‘yes’. But if you consider both metrics, the answer would probably be ‘no’.

How to avoid this?

1. Consider all important metrics when A/B testing. At Booking.com health metrics are by default added to A/B tests to avoid confirmation bias.
2. Let team members review A/B tests before going live.
3. Work with diverse teams. The more diverse the team, the easier it is to recognise biases. 🤝

Closing thoughts

We barely covered the tip of the iceberg, but I hope highlighting these biases can help product managers become more mindful of the effect these can have on their product. Thanks for reading!