The Science Behind Consumer Market Research
Recently CSA Research launched the third edition of our frequently cited “Can’t Read, Won’t Buy” research. We originated this topic of research in 2006 and continue to provide companies worldwide with the reliable data they need to plan and execute effective international growth strategies.
When you encounter data like this from us or any other source, put on your data scientist hat and ask these two questions that will tell you whether you can trust what you’re reading:
- What is the population? Population simply means, “Who is included in the group that we want to know about?” For these studies, we defined at the beginning that they would be people between the ages of 18 and 64 in 29 countries who use and/or purchase products/services online or by mobile app. To achieve our study’s goals, these are the people we needed to hear from.
- Is it a representative sample? A sample is the part or slice of the population that you acquired information about. It’s the data you have to work with. What’s most important about the sample is that it’s representative. In other words, the group you are looking at (the sample) has the same pre-defined characteristics in the same proportions as the entire group does (the population). When it does, it allows you to say with confidence that the results are true not only for those you surveyed but also for the full group. This simple principle is at the core of reliable statistical analysis. But acquiring a representative sample is often far from simple.
It is extremely difficult to reach thousands of people in 29 countries, in their native languages, that have the exact characteristics listed above. For our “Can’t Read, Won’t Buy” research series, we hired the multinational data research company Kantar and its network of panels that is the largest in the industry: 88 million research-ready respondents with 4,800 profiling attributes. To undertake a global study on consumers and businesses of this magnitude and provide the market with reliable data, partnering with a company like Kantar is mandatory.
We took demographic data for people in all 29 countries who use and/or purchase products/services online or by app and calculated the number of respondents we needed in each country, age group, and gender so that the sample would be representative. But the vetting didn’t stop at whether they fit the criteria we set – respondents also had to be answering truthfully and paying attention to their answers.
To achieve this, Kantar used its proprietary Honesty Detector software along with filters and trap questions to screen and validate every single response to both surveys. This combination winnowed a total of 31,933 surveyed consumers down to a representative sample of 8,709 in 29 countries. For the business study, the software vetted 6,703 business users to yield a representative sample of 956 in 24 countries.
Imagine if we had based our analyses on those nearly 30,000 rejected responses, or even a portion of them. The results and conclusions would be different, unreliable, and they wouldn’t apply to the millions of other consumers worldwide who didn’t take the survey. You wouldn’t want to make any decisions based on such results.
Can Answer, Can Trust
If you can answer those two questions and the answers make sense for the data, you can trust it. But if, for example, it’s unclear who was surveyed or specifically why they were included and others excluded, be wary.
Data matters, in particular during times of uncertainty. At CSA Research our mission is to properly identify both the goal of the research project and the population before acquiring a representative sample either on our own or with the help of a trusted data partner. It is neither easy nor quick to undertake research in this way, but it is the reason the core value of our market research brand is integrity.
Check “10 Facts About Can't Read, Won't Buy 2020,” as well as all of our Global Growth Research Series.
For additional, custom data cuts on consumers and businesses worldwide, contact us at firstname.lastname@example.org.
About the Author