Confirmation bias is all about selectivity: selectivity in the data that you pay attention to, and how you processes that data.
This cognitive bias has gotten a lot of attention recently in terms of how we consume and share news in the world of social media. The effect is stronger for emotionally charged issues and for deeply entrenched beliefs, which can be particularly dangerous in our current political climate. (Here’s an article that offers a great analysis of confirmation bias in the 2016 US election.)
What confirmation bias looks like in practice
Beyond news and social media, confirmation bias shows up a lot in advertising. McDonald’s ads have been telling people, “You deserve a break today,” since the 1970’s. This will resonate with those that think, “yes, I do,” which is a common belief (and why the ad is effective.)
How can you use confirmation bias to improve your designs
Take a side
A lot of websites and apps, especially those of big companies, try to offer something for everyone. They avoid taking a firm position on what users should do or think. But those who do take a position can benefit from the confirmation bias of people who DO agree with that position.
AHM Health Insurance in Australia has a great campaign that highlights how much people hate health insurance. Their billboards and digital ads admit that health insurance is annoying. This attracts and appeals to people who have that opinion. They might notice this ad more than more generic ads, because it confirm their beliefs that health insurance is annoying.
Yes, you will likely not be attracting those who don’t share those beliefs. You might not stand out to those whose opinions are not confirmed by your statements. But for those that do, you’ll attract more easily and retain longer.
Customer research is key in this approach. The only way you can speak to how your customers feel, in order to make statements that confirm their beliefs, is to speak to them. Qualitative research is the best tool, using interviews or remote user testing.
Personalise site content
If your goal is to get users to find and read your content, use this natural tendency to seek confirming information to layout and prioritise your content for each user. As I alluded to above, this tactic can be viewed negatively, depending on the content. In politics, helping users confirm their positions can lead to the large divides as we’re seeing in the US.
A study of 376 million English-speaking Facebook users found that users typically interact with a small number of news outlets and a limited set of pages. These behaviors allow news consumption on Facebook to be dominated by selective exposure, meaning that people are most often exposed to news sources that reinforce their existing opinions.
In other contexts, it can improve the user experience.
If you know a visitors opinion or desire, direct them to the landing page or prioritize site content for them. You might know their opinion because of the specific ad or article they clicked on, or based on their previous visit.
For example, say I click on a headline on social media that reads “Travel is the best way to spend your money” (something I would do because I want to read evidence to confirm my belief in that statement). The website that I visit would then know quite a bit about the content that I might notice and be more likely to enjoy, share, and recall. Other stories about travel and experiences would confirm my views and preferences, whereas articles trying to convince me to buy things might not. The site could surface the articles that relate to this belief. Not just related articles about travel in general, but articles about spending and saving money according to a value prioritising travel.
This is different than just recommending related articles, or articles the visitor might like.
This is understanding a visitors’ world view, and showing statements they agree with to affirm that your site or business agrees with them.
Research-based journey mapping is the UX tool that is key in this approach. Create a qualitative research plan to interview users, understand their motivations and behaviours, and deliver a personalised experience based on their beliefs. (I created a course about getting started with qualitative research, if you’re new to it.)