Make sure to familiarize yourself with the Audiences help doc as the foundation for this help doc. And also check out the Targeting Audiences help doc to see a slightly different function of this feature.
Collecting Data
The first step here is to start collecting data by publishing a survey. Go to the Surveys page and check that there is a "Live" on at least one post-purchase survey. If not, refer back to this article on how to set up your first survey!
Observe Reports
Make sure that you created the audiences that you want to use as filters before this step! More information on that here.
Next, head over to the survey that you're interested in to utilize the Filtering Audiences to analyze the data that you have collected! All you have to do is click into it from that Surveys page.
Select the date range and the audience from the drop down to start filtering!
Filtering audiences will work correctly regardless of what chart type you use to view your data (default bar graph, Attribution Report, Pie Chart, or Trend Report), so feel free to click the Chart Settings cog wheel and view your favorite chart.
Single Question -> Multiple Audiences
Use Filtering Audiences that use the same Audience Type to compare different segments on one question across the full dataset.
For instance, you can be asking the 'How did you first hear about us?' question and another "How old are you?" question in your survey. Seeing the results for the HDYFHAU question will likely change if you go from no filter, to applying Survey Question asked "How old are you?" Answer is "18-24 Age Range" Audience.
Single Audience -> Multiple Questions
Another way to utilize Filtering Audiences is to do a deep dive on a particular segment of your customer population within a single survey.
For instance, to analyze the segment of customers who discovered your brand on TikTok, set the Audience to TikTok Discovery (audience built by a TikTok answer to the HDYHAU question) and analyze each of the questions on your survey, filtering through the questions and toggling back and forth between Discovery = TikTok and No Audience.
Another great use case for Filtering Audiences is to overlay things like Product ID, Lifetime order count (new, returning, or high return for example)...etc on your existing results. Let's say that you launched a new product and advertised it heavily on Meta. Your results for the HDYFHAU results will likely vary if you overlay what the customers purchased in that case.
If you have any questions about Filtering Audiences or Audiences in general, please contact hello@knocommerce.com, and we'll be happy to answer whatever you need.