So much data is available. And, sophisticated AI (artificial intelligence) techniques can now help make sense of it all. So, data is taking an increasingly prominent role. As this unfolds, however, it is important not to lose sight of what the data really means.

Data can bring valuable insights about how to be more successful. Understanding the dynamics of a business is generally a crucial component of success. Data contributes to such an understanding. Experience also contributes to such an understanding. However, in a sense, experience can be viewed as employees accumulating extensive data about the business without necessarily formally recording it.

In today’s tech oriented world, modern approaches like AI seem to be in the limelight. But, older techniques can still play a valuable role. A reminder of this comes from a recent Wall Street Journal article, “Don’t Give Up on Focus Groups“ by Kate Murphy, which appeared May 24, 2021.

The article points out that big data “has seduced many companies into thinking they know their customers better than their customers know themselves,” however, “there is a growing realization that even if the numbers don’t lie, they can be seriously misleading. To really understand the beliefs, motivations and passions that move people, it is still necessary to sit down and listen to them, which is what qualitative research is all about.” The article explains that qualitative research includes “home visits, one-on-one interviews, and focus groups”. The article discusses Lego, which relied on data analytics and made marketing mistakes that turned around when the company shifted to doing qualitative research.

As I see it, the article’s advice to use qualitative research and listen to customers is valuable. After all, big data is not always the answer and AI can be overhyped. However, the article’s emphasis on talking to customers may or may not apply, depending on the situation. For example, the concept of knowing your customers better than they know themselves was advocated by the late Steve Jobs. Jobs headed Apple when it successfully introduced the iPhone without talking to customers and became the most valuable company in the world. Jobs’ viewpoint and Apple’s tremendous success may seem to contradict what the Wall Street Journal article is saying about the importance of listening to customers by doing qualitative research.

But, in my view, the key point here is that it is important to understand what the data really means. The Wall Street Journal is right that numbers can be misleading. That’s why it is so important to think through where the numbers came from, what they really mean, as well as whether and how the information might be influenced by its source. In Lego’s case, there were missteps because the company apparently did not really understand what its data meant until it talked with its customers.

On the other hand, Apple’s situation was far different than Lego’s. First of all, Apple had many staffers who were in a similar demographic as potential iPhone customers. Thus, Apple had internal capability to understand its customers, enabling it to possibly know its customers better than those customers know themselves. Furthermore, since the iPhone was quite innovative, many potential customers probably did not yet understand the innovation and still did not know that they would ultimately want it.

As Apple’s situation illustrates, an understanding of the customer does not always have to come from formal market research. Besides staffers who are similar to potential customers, another possible source can be employees who have extensive customer contact. Granted, these employees won’t know everything about their customers, so care must be taken not to over rely on them for what they may not know. However, they may have enough customer knowledge to provide valuable input for the business. For example, customer contact employees might be a resource when a company tries to get better results from big data by adding the human element to its AI (artificial intelligence).

The point of this is that companies must understand their markets and their business. If they choose not to talk to their customers via market research, they must judiciously evaluate whether they know enough about their market to skip that step. And, their assessment must be realistic. In other words, success is unlikely to come from merely trying to outguess customers who are deemed less knowledgeable about their needs than the company is. Instead, there must be good reasons why the company would know better than the customers. Otherwise, the deck can be stacked against companies that do not talk to their customers.

Furthermore, Lego’s experience doesn’t mean toy companies cannot do some of what seemed to work for Apple. In the past, the toy industry has successfully tested products on its people’s own children, particularly when the company was small and lacked the budget for formal market research. Again, the key point here is to evaluate the data and where it is coming from in terms of how good a source that data is for any particular situation. It’s not one size fits all.

Finally, even when data comes from asking customers, it is still important to think through what the data really means. Data collected by asking customers is not always free from the kind of bias that AI has been prone to. Customers who participate and respond to questions may not always be representative of the broader customer base. I wrote about this back in the 1990s in a peer reviewed article I authored in Marketing Research: A Magazine of Management and Applications. In that article, I told of a situation where the heaviest users were most likely to respond when questions were asked of customers via surveys. Consequently, in this case, due to bias in the survey’s representation, a customer database could more accurately estimate average usage by the entire customer base than did asking customers how often they buy.

So, in conclusion, regardless of whether using the latest AI techniques or talking with customers, it is essential to think through what the data really means. It’s important to look for consistencies across sources, and to think through whether and where the data can potentially be misleading. Doing so can help bring about more meaningful insights.