How Confirmation Bias Is Destroying Your Product

Probably the most harmful phrases in product development are: “Our customers will love this.” I’ve heard this declaration in numerous product conferences, normally adopted by months of engineering work and ending with the quiet disappointment of underwhelming person adoption. The perpetrator? Affirmation bias — our mind’s maddening tendency to hunt out info that helps what we already imagine.

As product managers, we’re employed to make selections. We analyze markets, collect necessities and prioritize options. The issue is, as soon as we have developed a speculation about what users want, we begin filtering all incoming info by way of that lens. Ambiguous suggestions will get interpreted as supportive. Unfavorable suggestions will get labeled as “edge circumstances.” And regularly, we assemble an alternate actuality the place our product selections are all the time good.

Associated: How Entrepreneurs Can Overcome Confirmation Bias

The person analysis theater

“Person analysis theater” refers to going by way of the motions of speaking to customers with out really being open to having your assumptions challenged. You would possibly acknowledge these signs in your group:

• Cherry-picking optimistic quotes from person classes whereas ignoring damaging patterns

• Asking main questions designed to elicit particular solutions

• Limiting your research to customers who already love your product

• Decoding silence or confusion as settlement

• Dismissing damaging suggestions as “they only do not get it but”

Look, I get it. You’ve got already advised your leaders and traders in regards to the superb characteristic roadmap. You’ve got employed engineers primarily based on sure technical assumptions. Your total firm narrative is perhaps constructed round a selected imaginative and prescient of what customers need. Altering course feels inconceivable.

However staying on the doomed course is worse.

Associated: Do You Know What Your Customers Want? Are You Sure?

Breaking the bias cycle

So, how can we really repair this? How can we create processes that challenge our cherished assumptions as an alternative of reinforcing them? Listed here are some sensible approaches I’ve seen work:

1. Separate knowledge assortment from interpretation

One workforce I labored with adopted a follow the place the folks conducting person interviews weren’t allowed to interpret the outcomes. They may solely doc precisely what was mentioned. A separate workforce — one with out emotional funding in particular outcomes — would then analyze the transcripts. This lowered the tendency to listen to what they needed to listen to throughout interviews.

This separation creates a wholesome stress. The interview workforce focuses on asking good questions slightly than main customers towards predetermined conclusions. The evaluation workforce spots patterns with out being influenced by customers’ tone or the interpersonal dynamics of the interview.

2. Actively search disconfirming proof

Make it somebody’s particular job to play devil’s advocate throughout analysis planning. This individual must be asking: “How would possibly we disprove our speculation?” slightly than “How can we validate our concept?”

For instance, as an alternative of asking “Would you utilize this characteristic?” attempt “What would forestall you from utilizing this characteristic?” The primary query virtually all the time will get a well mannered “sure.” The second provides you precise obstacles you may want to beat.

3. Take note of conduct, not simply opinions

Customers are notoriously unhealthy at predicting their very own future conduct. They will enthusiastically inform you they’d positively use your new characteristic, however when it launches, they follow their outdated habits.

I’ve discovered it rather more invaluable to look at what customers really do slightly than what they are saying they will do. This implies analyzing usage data from current options, creating prototype experiences the place customers can exhibit preferences by way of actions, and conducting discipline research the place you watch customers of their pure surroundings.

4. Create a tradition that rewards altering course

In case your workforce will get punished for admitting they have been improper, guess what? They are going to double down on unhealthy concepts slightly than acknowledge the necessity to pivot.

Good firms construct ceremonies that commemorate studying and adjustment. Some startups have achieved “Pivot Events” — precise celebrations when the workforce made a significant course correction primarily based on person insights. They actually popped champagne once they killed options that analysis confirmed would not succeed. This despatched a strong message: Studying is valued over cussed persistence.

5. Diversify your analysis individuals

For those who solely speak to your most enthusiastic customers, you are creating an echo chamber. Be sure your analysis contains:

• Potential customers who selected competitor merchandise

• Former customers who deserted your product

• Present customers who hardly ever interact together with your product

• Customers from completely different demographics and use circumstances

This variety helps expose blind spots in your understanding.

Associated: 3 Cognitive Pitfalls That Are Ruining Your Business — How to Unravel the Biases in Decision-Making

The paradox of experience

Here is the painful fact: The extra skilled you might be in your area, the extra vulnerable you turn into to affirmation bias. You’ve got seen patterns earlier than. You’ve got developed instinct. Typically that is extremely invaluable. Different occasions, it makes you dangerously overconfident.

The answer is not to disregard your expertise. It is to pair your hard-earned instinct with rigorous processes that take a look at your assumptions. One of the best product leaders I do know have robust convictions loosely held. They make daring bets primarily based on their experience, however they’re fast to regulate when proof contradicts their preliminary hypotheses.

Ultimately, the market does not care about your good imaginative and prescient or your elegant answer. It solely cares should you’ve solved a real problem in a means that matches into customers’ lives. And the one method to know that for positive is to consistently problem what you assume you understand about your customers.