Some say social media can work for data science ...
You are asking a great question and there is a proven solution. Internet has opened up a huge world for us, where diversity prosper. One of the biggest hidden secret for Survature users is - as long as you can find a place where people sharing similar interests congregate on social media, Survature can get you their psychology in a matter of minutes.
While that promise sounds too good to be true and you haven't see anyone deliver that, most people haven't seen the mechanical part of that solution yet. Survature users have that at your finger tips, because our data-science user-support provides the full solution.
Part 1 - Have a Project Landing Page
You might not have thought of that as necessary. You might not have thought of that as feasible. You might not have thought of that as effective. You might not have thought of that as quick to do. Regardless of what the reason, you haven't done it before. Guess what, that's the first secret ingredient. Even better, Survature's data-science user-support can do all of that for you.
For example, the following is a screen capture of an actual project landing page that we created for our users at the Global Supply Chain Institute, the University of Tennessee. This landing page is embedded with the AI required to manage target logics, etc.
With a landing page, or a set of carefully designed landing pages, you get the ability to make very appealing social posts in the places where your target audiences congregate. In addition, our user-support knows the best practices to optimize social posts as well as emails. Please do not overlook the appearance. The appearance is your brand. There is no engagement, until you have gotten the brand right.
Part 2 - Test, Improve, Test, and Improve Again
As you can tell from all marketing, nothing works right out of the gate. Anything that becomes a case study has been through repeated failures. Thankfully, with knowledge of data science and the ability of today's technology ecosystem, refinement iterations can be as fast as on a daily cycle.
Would you be surprised that one of the most successful projects we have seen - it's a customer-research project for an out-of-this-world kind of product, our user in the end tested with 15 LinkedIn groups and 20+ Facebook groups. They got amazing data products and complete profiles of their potential customers, didn't pay a single dime for a "sample" (i.e. those "professional survey takers"). The overall process of adding to the dataset while making continuous improvements took about 2 weeks.
Part 3 - Separate / Aggregate by Source
Throughout the whole process, it's crucial to use targeting to distinguish each source and each version in the test-improve cycles. The corresponding help-doc for targeting is here.