Randomized (aka general population) audience
If you aren't sure who your target market is yet, or you're trying to establish a baseline test to drive your creative workflow, we suggest starting with a larger general population panel. What this means is that anyone on our panel can answer- no matter their demographic or behavioral segmentation.
It is first-come-first-served based on respondent availability so it will be a random sample of people. However, you can add demographic questions in the final step of the poll builder to report on traits to help you understand which way your product/service may skew based on consumer sentiment and moving forward, segment accordingly.
Targeted audience
For those of you that know your product skews towards a certain segment of people, then choose those traits when building your poll to make sure you only reach the people you want to hear feedback from.
Oftentimes we see people use a lot of their budget towards stacking traits and choosing a niche or very narrowed audience when it's absolutely not necessary. If these are the only people that will be purchasing or interacting with your brand/product, then that is perfect, but if you are choosing blindly based on a hunch or assumption, you'll likely end up wasting those funds that could be better allocated elsewhere. If unsure, always start broader and narrow down as you collect consumer data.
The more requirements you set in order for respondents to partake, the more expensive the poll will be and the longer it will take to assemble and complete so just keep that in mind when planning your tests.