The first step for a creating a polling (text survey or IVR survey) project is to create a survey sample list. To read more about text vs. IVR surveys, check out this article.
💡If you are interested in adding a randomized voter sample within the list builder, please reach out to Numinar staff for assistance.
Step 1/7: Click on Lists
Step 2/7: Click on New List
Step 3/7: Label and create list
Step 4/7: Select filters for an IVR Poll List Example
IVR poll lists should have the following filters: "has land line" w/true and "do not call" w/no data. You can use the reliability filter if you'd like and the random sample to achieve a desired number of voters.
Step 5/7: Select filters for a Text Survey List Example
Text polling lists should have the following filters: "Has Cell Phone" assigned true and "Do not Text" assigned No Data. You can use the reliability filter if you'd like and the random sample to achieve a desired number of voters.
Step 6/7: Click get total counts to check your list size
If running a poll, typically the list universe should be around 20k voters. However, you can adjust as you see fit! We estimate a 1-2.5% response rate (give or take). Sampling 20k voters should yield an ideal number of response to be used for a model (n = 500) if that is your use case.
Step 7/7: Click on Save
💬Frequently Asked Questions
How many voters should I have in my random sample list?
How many voters should I have in my random sample list?
If running a poll, typically the list universe should be around 20k voters. However, you can adjust as you see fit! We estimate a 2.5% response rate (give or take). Sampling 20k voters should yield an ideal number of response to be used for a model (n = 500) if that is your use case.
Do I need to use the "do not text" and "do not call" filter?
Do I need to use the "do not text" and "do not call" filter?
Yes! Even if it's your first outreach project, it's a good idea to use those filters and select "no data" value to ensure you aren't doing outreach or polling to voters who have unsubscribed.