A company wants to understand purchasing power in a large market, so it conducts a survey and asks hundreds of people about their income and spending. Everyone answers honestly, the data is collected correctly, and the averages are calculated precisely.

And yet, the result may be wrong.

Not because anyone lied or made a calculation error, but because the company asked the wrong people.

Who you ask may matter more than what you ask

In a Marketing Management course at Al Yamamah University, Dr. @Yosra Missaoui discussed an example that illustrates the problem of sampling bias: if we want to estimate the average income of Saudis and collect data from people in a VIP lounge, the issue isn't with their answers, but with how well they represent the population we're trying to understand.

Here's the paradox: the data can be accurate, while the conclusion drawn from it is misleading.

A sample is the group from which a researcher collects data to understand a larger population. If that group is skewed toward a particular segment from the outset, its characteristics may carry over into the results as if they represented the entire market.

Imagine a clothing store that wants to find out how much consumers are willing to pay for a new product, but surveys only customers at one of its branches in a high-income area. Their answers may be perfectly accurate, but on their own, they aren't enough to describe how willing consumers in the broader market are to pay.

If a pricing decision is based on this result, the problem may have begun long before the price was set—with the choice of sample itself.

More data won't fix bias

The obvious solution might seem to be increasing the number of participants. But a thousand responses from an unrepresentative group won't automatically become an accurate picture of the market just because the number is larger.

This explains why the strength of market research isn't measured by the amount of data alone. The more important question is: Is the data collected suited to the question we're trying to answer?

For companies, this can affect pricing decisions, product development, market selection, and marketing campaigns. A decision may appear to be backed by numbers, while the problem lies in how those numbers made their way to the company.

In a world where collecting data is easier than ever, the real challenge may not be getting more of it, but making sure we're listening to the people who represent the market we want to understand.

“Before reading the results of any survey, the question isn't just what did the participants say? It's also: who are the participants in the first place?”