What is crosstabulation?
'Cross' means to link different things together. In other words, cross analysis is to link and analyze multiple data. This is because when you link the response results of one variable with other variables, embedded data, and respondent profile information, you can gain richer insights.
For example, you can cross-analyze variables that ask about brand preference by gender and age. Then, you can compare and analyze the differences in brand preference between women and men, as well as preferences by age group: teens, twenties, thirties, and forties. Then, you can easily identify which segments have particularly high or low brand preference, and through this, you can establish sharp actions and strategies for specific target customers.
Understanding the ‘Analysis Object’ and the ‘Analysis Unit’
To utilize crosstabulation, you need to know what a crosstabulation table consists of. A crosstabulation table consists of an analysis subject and an analysis unit.
Analysis Subject: This literally means the subject of analysis or the data to be interpreted. Based on the table below, the analysis subject is 'Preferred Wireless Earphone Brand'.
Analysis Unit: This refers to the unit by which the analysis target is divided or the data that serves as the basis for interpretation. In the table below, the 5-year age unit is set as the analysis unit. This allows you to analyze the response results for preferred wireless earphone brands by age.
Analyzing crosstabulation tables in the Crosstabulation tab of Analysis
Let's look at how to cross-analyze the 'Usually Preferred Wireless Earphone Brand' in the example image we looked at earlier by age in 5-year increments using the Crosstab tab of the Analysis.
① Understand the overall trend by 5-year age group
If you sort the responses in the cross-tabulation table, the vertical axis will change to the most preferred brand order. If you look at it, you can see that the order of preference is Apple > Samsung Galaxy Buds > Bang & Olufsen.
Let’s see if the tendency based on the entire respondents is reflected in the age-based response results. The percentage of those aged 15–19 and 20–24 who prefer Apple is 80.3% and 75.9%, respectively, which is considerably higher than the overall average (average: 65.5%). Meanwhile, the percentage of those aged 30–49 who prefer Bang & Olufsen is over 20%, which is higher than the average (average: 17.7%). If we divide it by age like this, we can see the temperature difference in the preferred brand by generation at a glance.
If we were Bang & Olufsen marketing managers, what insights could we gain from this? Bang & Olufsen is a high-priced premium wireless earphone brand and is relatively popular among people in their 30s and above, so we can conclude that we should focus our marketing efforts on people in their 30s and above.
② Analyze data in more detail by changing the analysis unit
As a Bang & Olufsen marketer, you could have concluded that you should focus on the target of those over 30. Are there any differences by gender? You can find out about this by adding gender to the analysis unit. For reference, in cross-analysis, you can stack up to 3 analysis units vertically.
When we look at it, men in their 30s show a relatively higher preference for Bang & Olufsen than women in their 30s (29.0% and 24.0%, respectively). In addition, we can see that this difference in preference between men and women is even greater in their 40s (37.4% and 18.4%, respectively).
The above analysis results show that Bang & Olufsen is relatively preferred by those in their 30s and above, and among them, it is more preferred by men. In this case, Bang & Olufsen marketers can use marketing strategies based on information about more specific target customers.
③ Use cross-analysis on other variable data as well
Cross-tabulation is not an analysis method that is limited to specific situations. If you have a variable that contains information you want to understand and a criterion for splitting that variable, you can create a cross-tabulation table that is suitable for that.
As in the example above, if we are a premium wireless earphone brand marketer, we can conduct a survey consisting of the following questionnaire, and all of these variable results can be analyzed.
Please select the earphone brand you are currently using. (Usage Status)
Are you willing to purchase a premium earphone brand in the future? (Purchase intention)
What is the most you are willing to pay for a pair of earphones? (Purchase Price)
What are the most important features to consider when buying earphones? (Key purchase considerations)
If you divide the response results of the above variables into the following analysis units, you can conduct a more detailed analysis. Therefore, it is a good idea to think in advance about which data will be the analysis target or analysis unit from the survey design stage.
Demographic variables such as gender and age
Behavioral variables such as whether you are a heavy user or light user, or your main purchasing channel
Need variables such as price sensitivity
If you haven't done your own research yet but would like to learn more about crosstabulation, please visit the sample survey link below. In the Crosstabulation tab, you can create various crosstabulations by directly utilizing the cards in the left sidebar.