As content websites fragment audiences into more and more tightly defined demographics, your site's target is by definition going to look less and less like a "representative sample" of anything.
That's not to say that a survey of U.S. consumers or U.S. internet users (still two different demos) isn't useful any more.
For instance, last week's survey by ExactTarget (@PennOlson) showing that active Twitter users are the most "influential" folks on the internet had some really intriguing data. The survey found:
- 72% publish blog posts at least monthly
- 70% comment on blogs
- 61% write at least one product review monthly
- 61% comment on news sites
- Men are more than twice as likely as women to follow brands on Twitter to interact with the company (29% compared to 13%).
If you're trying to make the case for why your brand should be on Twitter, smartly, this data is good fodder. Couple that with Nielsen's BlogPulse data that there are nearly 145 million blogs generating 782,000 new posts in the last 24 hours and you get a sense of scale for how much content is being churned out every day. This is important for any publisher or marketer to understand. While any audience is likely to look something like this, your specific audience might only match in some ways.
An interesting wrinkle, however, is what Sarah Evans of Chicago-based Sevans Strategy did with the ExactTarget data. Recognizing that her audience is not "representative," she asked her readers to take a similar survey. That kind of data is a long way from scientific. You're likely to get your most-engaged audience, but even that isn't necessarily your real target audience.
But that doesn't mean it's worthless. It could potentially bring some hidden facts to light. And it's not hard to accomplish. With tools like Survey Monkey or Twitter-based TwtPoll, you can put together a survey in no time and see how results of a national or global survey compare to the most important demographic: yours.