You collected your survey answers. Now what? Learn how to analyze your survey data and give meaning to your results.
You carefully thought about your survey: you created the questions, set up your questionnaire and you collected results. Now the real fun begins: it is time to understand what your results actually mean, in other words, it is time to analyze survey data.
Data without analysis do not bring many insights. By analysing your results, you can draw conclusions and gain insights that can help you make decisions and improve your business.
A sound analysis of your survey data is crucial because it can influence the strategy of your business. When the analysis is well conducted, then it can shed light on new opportunities for improvement. On the other hand, an erratic survey analysis can lead to the wrong business decisions.
To draw meaningful conclusions from your survey, you first need to check the number of respondents. This number will tell you how large your sample size is so how much you can rely on the data you have gathered.
Depending on the survey tool you are using, this could already notify you if your sample size is not large enough. For example, in the UserReport dashboard, if the sample size is too small, you will see a red “n” symbol next to the results you have collected. Until the red “n” is there, you will know you need to wait before you can analyze survey data: easy enough!
As a rule of thumb, UserReport considers results above 50 to be a sufficient sample size but this, of course, depends on the segment of people you want to analyse with your survey.
If for example you’re running your survey on your website and you aim to know more about your users, then the sample size will depend on the number of website visitors you have. Do you have 1 thousand visitors per month? Or 100 thousand? The sufficient sample size will change a lot depending on this.
What does aggregating numbers mean? It is easy: you need to count how many people answered one question and how many chose each answer option.
Say you are asking the participants how many people live in their household. First, you need to know how many people answered your survey, then you’ll need to count the total number of people who chose each answer option.
Again, if you are using a survey tool, you do not need to aggregate numbers: you will find these data in the reporting interface. Here you can see how UserReport shows aggregated data:
Get beautiful reports with UserReport free tool
Once you have aggregated numbers, it is time to give some meaning to them. It is generally easier to look at percentages than whole numbers and also to compare them. It is also good practice to start from the quantitative data you have collected and leave quantitative for later.
Again, most likely your survey tool will show you the answers to your questions also in percentages forms (see picture above). Thanks to percentages you will quickly have an overall idea of which answers were the most popular and this will make it way easier to analyze your survey data.
This is part of survey data analysis when you really start drawing conclusions about your data. Cross tabulating can help you group your respondents into different groups and add some background to your data.
When you were creating your survey questions, you probably thought of the different comparisons you would have found out. For example, let’s say you expect men who practice sport to also have kids. You will first need to look at data for your first question:
How you practice sports at least once a week?
You are interested to add an extra layer and understand if sportsmen also have children. So you will cross-tabulate the previous question with this one “ Do you have children in your household?” A cross-tabulation (or crosstab) is a table that records the number (or frequency) of respondents that have the specific characteristics described in the cells of the table.
To do this you can export your data and use a Google sheet. That said, there is even a faster way. If you are using UserReport, you can build customer personas based on specific answers and easily segment your respondents. Check out this video to see how:
Giving meaning to your survey data is hard. Sure, percentages can help you analyze your survey data but it is still hard to understand what can be considered a “good” result. For this reason, you should have benchmarks: data you can take as reference point when analyzing your survey data.
The easiest way to have benchmarks is comparing the data you have gathered with the data you collected with the previous survey. Obviously, in order to do so, you need to ask the same questions every time you conduct the survey.
Month by month comparison is great and can reveal you many insights, also it can show if there are any trends.
As always, you could do this manually but better use a survey tool that allows you to compare time periods and show which answers changed and how much:
Last but not least, it is time to talk about how to showcase the data you collected. One thing you should bear in mind is that bare data are boring. In fact, it is hard to grab someone’s attention just by listing percentages. You should try to present your result with a narrative, with a story.
You should then always provide a comparison maybe with the data you collected the month before, this way your audience will know whether results are positive or negative. Illustrate the correlation between two variables and how one clearly influences the other.
Finally, give your audience the information that is relevant to them, and that forms a solid basis for making decisions.
Of course, layout plays its part in this last phase. Use graphs and colours to highlight similarities and differences and if you are using a survey tool with a good reporting interface design, make the best out of it!
UserReport is a completely free survey tool that comes with beautifully designed reports where you can easily give meaning to the data you collect. Try it for free today!