For anyone in market, social, and opinion research, Twitter polls might seem to be pointless, and a waste of time. There are no controls around the sampling methodology and it is convenience sampling at its worst. This is one case where calling it accidental sampling seems more appropriate.
So here’s the situation. You have 5 great products and you’re wondering if you should build a sixth one. What would happen if you built the sixth one, and people stopped buying your other products?
These days, it seems as though every survey needs to be designed twice. Once for regular desktop computers and then a second whereby we redesign the original survey to work better on a smaller mobile device like a table or phone.
What if I told you someone was giving out free pizza to anyone whose favorite color was orange? What if someone was giving out free iPhones to anyone who believes voting is important? What if someone was giving out brand new big screen TVs to people who were 5 foot 11 or taller?
I’ve been in the marketing research business long enough to hear the stories. A brand management team of 25 year olds working on developing new types of walking aides like canes and walkers. Women working on moustache trimming devices. All male speaking panels discussing a research project about tampons and pads.