Whether you’re conscious of it or not, your own internal biases will affect how you perform market research.
When you’re creating a survey for your audience, your hidden biases may change how you word questions or answers, which can result in misleading data. Worse – your company might act on this data without any awareness that something has gone wrong.
That’s the major problem with bias – it’s not that it happens, it’s that you might not be aware that it’s happening until it’s too late.
The solution is to eliminate as many biases as possible at the start, before you send out your survey. While it sounds simple, it’s not easy – how do you eliminate something you’re not aware that you have? In order to give yourself the best chance to conduct accurate market research, you should consider the four different types of bias below and implement strategies to eliminate them.
Removing Confirmation Bias from Market Research
This is the big one – it’s the most common type of bias, and it affects us all in various aspects of our lives. In short, confirmation bias is seeing only what you want to see.
For example, if you believe that left-handed people are more creative, and you meet someone who is both left-handed and creative, you’ll likely point to this as evidence to support your claim. However, if you meet someone who is left handed but not creative, you’ll dismiss them as an outlier.
When you have a confirmation bias, you only focus on evidence that supports what you already believe.
Obviously, this becomes a problem when you’re conducting market research. If you believe that potential customers are most interested in a particular feature, you might create a survey with an inordinate amount of focus spent on that feature. Then you’ll point out how many people are talking about that feature – even though you didn’t give them anything else to talk about!
In order to prevent confirmation bias, you need to constantly be re-evaluating both the questions you ask and your interpretation of the results. Ask yourself: when it comes to your market research, what do you want to be true? Perhaps you want to believe that people come to your website for its long-form articles. Now that you know what you want to be true, you need to use your market research to poke holes in that desire. For example, rather than focusing your survey around improving long-form content, you would straight up ask your audience: why do you come to our website?
Managing Question Order Bias in Market Research Studies
When you ask someone a question, you put an idea in their head. These ideas can influence their answers to subsequent questions. As an example, look at the next two questions:
- What is your favorite thing about elephants?
- Name an animal that is grey.
While this example is a bit obvious, you can see the question order bias in action.
Essentially, when you ask questions in a certain order, you put ideas into the heads of your respondents and lead them down a particular path. When asked to name an animal that’s grey, your first response may very well have been “elephant,” but by asking you about elephants in the preceding question, the odds have increased astronomically.
Most often, market researchers will slip up when creating surveys that ask their audience to rate products. People tend to rate products in a relative manner – if you have two products, people will rate the second product based on how it compares to the first, rather than how good they think it is objectively.
The solution is to ask general questions before you ask specific questions. Also, be aware of when you ask people to rate things, and understand that they’re likely rating things subjectively based on what they’ve already seen in your survey.
Identifying and Eliminating Culture Bias in Market Research Studies
People have a tendency to assume that everyone sees the world the same way and hold similar values. This is culture bias.
As a market researcher, you can never make the assumption that your respondents hold the same values as you.
Culture bias crops up in how questions and answers are phrased. For example, let’s say you’re creating a survey to test what kind of new pizzas people might be interested in. You and everyone you know loves pizza that has pepperoni on it and hates pizza that has shrimp on it. You might innocently phrase your question like this:
What topping do you prefer on pizza: delicious pepperoni or shrimp?
Due to your culture bias, you’ve attached a positive value to one of the answers – the answer you subconsciously believe is “correct.”
The best way to remove this bias is to show unconditional positive regard for any answer a respondent might make.
Keep in mind that culture bias is, in many cases, impossible to remove completely – but you need to be aware of how you see the world, and the assumptions you’re making at each stage of the market research process.
Removing the Most Dangerous Bias from Market Research: Irrational Escalation
While not as common as the others, this bias is one of the most dangerous. Irrational escalation bias is similar to confirmation bias. The main difference is that irrational escalation bias happens after you’ve already made a financial commitment. It’s a form of sunk debt.
Irrational escalation bias occurs when you’ve produced a product and put it up for sale – only now there’s new research that shows customers are probably not interested in the product. Rather than cutting your losses and accepting you made a mistake, you double-down and convince yourself that the research is wrong, that if you just push your product a little harder, you’ll find success.
The solution is to be open to what your numbers are telling you. If you have real world results, listen to what they say. Ask yourself: if you hadn’t already launched the product, with the information you’re now receiving, would you still launch?
100% is Impossible, But Don’t Stop Striving for Bias-Free Market Research
Even the best market researchers will never fully eliminate all biases. But you need to be aware so you don’t inadvertently send yourself, and your business, down the wrong path.