Imagine making an important strategic decision based on data. You open your dashboard, check the numbers, and — hurrah! — the numbers confirm what you already thought. The market is picking up, customers are happy, and your investment choice seems brilliant. But wait a minute... is that really true? Or did your brain just tricked you?
Welcome to the wonderful world of confirmation bias, the inner cheerleader of your beliefs.
What is confirmation bias?
Confirmation bias is the tendency of our brain to primarily seek, remember or interpret information that confirms our beliefs. It's like typing in a Google search to confirm you idea. And let's face it: who ever clicked on the second page of Google when the first one says exactly what you wanted to hear?
A BI dashboard that only shows your successes… that’s not business intelligence — that’s just sweet-talking your numbers.
This bias isn't a stupid mistake, but a smart (if sometimes misleading) trick of the brain: it saves energy by ignoring the unknown and embracing the familiar. Unfortunately, in a world of data and dashboards, that can lead to expensive fallacies.
How does it work in practice?
Suppose you are a sales manager and you notice that the number of quotes won is declining. You suspect that the price is too high and dive into the numbers. And yes: “price” is the reason for dropping out. Case closed, right?
Not so fast. You could just as easily have looked at the response time of quotes (slow as a snail?) , the customer experience (an automated mail that sounds like it was written by a printer?), or supply quality (as predicted by your auntie Barbara and her crystal ball). But because you already had an idea, you were just looking for confirmation. And so you may be making adjustments in the wrong direction. Or in management terms: you're driving 120km per hour... but on the wrong exit.
Why is confirmation bias so insidious in a business context?
In organizations, confirmation bias is a real magician: it makes decisions look like objective insights, while they are actually heavily influenced. Especially in companies where “data-driven” working is a badge of honor, it is tempting to think that numbers tell the truth. But data doesn't lie... unless you whisper it to her.
Confirmation bias occurs in three flavors:
1. Individual level
An analyst who believes customer satisfaction is top notch rarely looks for disgruntled responses. Why bother yourself with complaints when you can just brag about straight A's?
2. Team Level
A sales team that swears by “price is the problem” will skillfully ignore any other signal, even if the customer actually complains about the help desk that sounds like they're still on dial-up internet.
3. Organizational level
Companies that see themselves as “disruptive” sometimes fall so in love with their own hype that they dismiss market signals such as “customer misunderstanding”. Until suddenly no one buys anymore. (Nokia is calling.)
Signs of confirmation bias in your company
Are you starting to suspect that there is something wrong with your data relationship? Note these signs:
- KPIs that always have a positive explanation. Even a decline is “an opportunity for improvement”. (As if a flat tire is especially useful for practicing your tire changing skills.)
- Anomaly in the data? It is “noise”. Read: hindering the success story.
- Exactly the same report every month. Drag, but in tables.
- Many sentences that start with: “We know that...” or “Our experience shows that...”
- Teams that talk to data but not to each other.
What does confirmation bias cost?
The damage caused by confirmation bias is difficult to measure exactly, but here are some of the problems:
- Wrong investments: You start a project on your gut with numbers as an excuse, and end up with a lot of extra inventory and a bitter budget.
- Unhappy customers: You rely on a high NPS score, but you don't notice that the complaints are piling up in your spam folder.
- Process inefficiency: If you only celebrate the successes, no one will care about the leak under the hood.
In short: a BI dashboard that only shows your successes… that’s not business intelligence — that’s just sweet-talking your numbers.
How to break confirmation bias
Time to take off the rose-colored glasses. Here are five ways to tame the bias:
1. Ask yourself the opposite question
Instead of asking, “How does this confirm my idea?” , you ask the question: “What could undermine this idea?” A bit like a job interview: you're hoping for a yes, but you're preparing for the question “What is your biggest weakness?” (Spoiler: Perfectionism doesn't count.)
2. Work with scenarios
Let dashboards show different realities. This way, you don't get tunnel vision, but a panoramic view. (And yes, that also means a view of the less beautiful landscape.)
You're better off with someone who dares to say, “Maybe we're wrong.”
3. Involve a “red team”
A team that is tasked with shooting holes in your analysis. Sounds annoying? Yes. Does it work? Absolutely. Think of it as a mandatory reality check, but with coffee.
4. Show anomalies
Let your dashboard show not only averages, but also outliers. One angry customer with a clear point is sometimes more valuable than a hundred anonymous thumbs up.
5. Compare data sources
Integrate your CRM, ERP, Excel schedules, and notes to post-its. When four systems tell four different stories, you know you don't have a truth — but a thriller with multiple storylines.
You can contradict data
In a well-functioning data-driven company, it's okay if the numbers contradict your gut feeling. On the contrary: that is often the starting point for growth.
Because those who are always right learn nothing. And let's be honest: even your GPS sometimes requires a recalculation. So why not you?
So dare to be critical. Not just for your data — but also for yourself.
To conclude
Confirmation bias will never disappear. It is an integral part of being human. But with some awareness, a good dose of self-reflection and the right tools, you can limit the damage and drastically improve the quality of your decisions.
So, the next time you think “see you,” ask yourself: do I see everything — or only what I want to see?
Your brain wants to be right... but your company is better off with someone who dares to say, “Maybe we're wrong.”
