facebook polls : speed vs statistical significance

As a bootstrapping entrepreneur, if you have $100, you would first give them to your lawyer, then to your accountant, and then lastly to your market researcher. Facebook launched polls recently, which enables the users to quickly and cheaply($5 basic insertion fee plus $0.10 per answer) set up a poll (a question with multiple possible answers) and target it to the Facebook community (based on some very high level criteria - age, sex, profile keyword, etc.)

I am impressed by the simple design of workflow of poll-creation, implementation and analysis. But, the most exciting part was the speed. Facebook claims that they can get answers as quick as within 30 minutes (if you pay $1 per answer), but even when I chose the cheapest option ($0.10), I got the answers fairly quickly (I set up the poll, went to sleep and the answers were ready by the time I woke up -- and I didn't oversleep!!)

I set up a sample poll with a simple question?

Why do you go to a conference?

 

I was happy with the answers that I received, but I am not confident about the statistical significance of the data. If I were to make a critical product design decision, could I rely blindly on results of Facebook polls? I don't know. Probably not, at least now.

Overall, I do believe that Facebook will become a major cultural and economic force, leading up to a huge value creation for its founders and investors. If you've tried their polls, please share your experience. My friend Robert Scoble had doubts about facebook's appeal earlier in December 2006, but he's expressed positive notes about Facebook recently. Facebook is rapidly creating a tightly bound network/community and it is intelligently leveraging its vibrant community for lucrative applications, besides advertising.

Michael Arrington from Techcrunch describes Facebook polls as a marketer's paradise and I agree.

get on Facebook.

-Kintan

Technorati tags: facebook , facebook polls , market research , entrepreneur , marketing