Data overlays are a way of segmenting your audience without requiring you to ask additional or sometimes intrusive questions. You can attach details that are already known about a survey participant (such as information about a returning customer) to the survey given to that individual. These details can be used during survey analysis to allow insights such as priority comparisons between customers based on how much the individuals purchased.
Use and formatting
To use data overlays, the first thing you’ll need is the link to your survey. This can be found from the account dashboard. For a running survey, click the Share button to obtain your survey’s public link. It will look something like this:
One you have your survey link, adding segment overlays is done by appending variables that you would like to track to the survey link that the survey participant will then use. For example,
In the above, the variables to be tracked for this survey participant are price and numOrders. The ranges are up to you. For example, priceRangeB could be $100–$150, and ordersCountA could be 0–5. You need to make sure the ranges you use strike a balance between having enough granularity and creating too much overhead during analysis. We recommend creating 5–10 ranges for any continuous variable.
It’s suggested that you obfuscate the variable names and ranges so that your customers opinions are not influenced.
Obfuscation means changing the names of each variable to something that you will understand but that is meaningless to a survey participant. The following rules apply when creating variable names for survey overlays:
For example, the first variable added to the URL could be in this format:
?name=value. The full link would be:
You can add as many variables to the URL as you wish. Add additional variables by joining them with an ampersand &. You should have something like this:
The variable names and range names can only use alphanumeric characters (A-Z, a-z and 0-9). Our overlay queries are not case sensitive, though. RangeA is treated the same as rangea by our system. After obfuscation, the above example could look like the following:
The order of the pairs is not significant.
If a customer doesn’t have a variable, leave the variable out of the link. For example, if it’s a customer that does not have a value for price, leave price out of the link for that customer’s URL. Do not leave a blank value, like this: