This has been done to death in the incentive based affiliate marketing world for some years now.
The quality of data that will be generated from these surveys is going to be extremely poor and I would be interested in seeing how they're able to sustain advertisers (market researchers).
I'll second this, we put together a system that did this around 2008 and while it worked out well initially within less than a year it became a loss for our company. We couldn't charge much for the data as even with large amounts of filters and detection techniques for bogus data and users the results were so bad no one was willing to pay for them any more. Within a year we scrapped the project and ripped it out of our ad system. People are more than happy to answer a survey to get to paid content, they are also more than happy to supply bogus answers to the questions.
That being said, I'll be interested to see what, if anything, makes this different than all of the surveys running in the affiliate marketing/lead generation space.
We agree that maintaining data quality is the biggest challenge and really the crux of this business. We have initial filters that look for incorrect answers to "checker" questions (e.g. What browser are you using? In which time zone do you live?), an outlier response speed (too slow or too fast), etc. And we're working on a more sophisticated system to really address this.
Data quality with initial partners has been good. We'll be working hard to keep it that way.
The quality of data that will be generated from these surveys is going to be extremely poor and I would be interested in seeing how they're able to sustain advertisers (market researchers).