This morning, I checked the Autos Insider page of (as I do every morning). I was greeted with the headline: Harris survey: 6 in 10 predict demise of terrestrial radio in 5 years. It went on to state that “Nearly 60 percent of Americans predict the death of terrestrial radio will happen within five years, according to results from a Harris Interactive survey released Tuesday.” 

Check that. An interactive survey, meaning it was done online.

Now, I have nothing against using online surveys! I’ve done many. As use of the internet has grown, online surveys have become more and more representative of the population as a whole, and have advantages over other methodologies. But I think this particular survey is a misuse of online research.

It’s not because I don’t like or agree with the results (though ditto on both). It’s not really even because the survey was done on behalf of Stitcher — an online “radio” service.

It’s a misuse of online research because it pits terrestrial radio versus online radio, and it was done online…among respondents who are likely heavy users of the ‘net, adept at using the ‘net, who likely spend more time online than average, who are more likely to already subscribe to Stitcher or Pandora, etc. than the population as a whole. More likely than not, they’re not representative when it comes to this particular issue.

Think about it. If stations went on the air and asked listeners to call in and vote on whether they prefer “regular radio” or online radio, do you think that would be representative of the population as a whole??? Of course not!!! I’d bet would turn the results of Harris’s survey on its ear. And it would be be at least as erroneous.

If you’ve visited this website before, you know that we’ve done online research among Pandora listeners. Online is a great place to find Pandora listeners, and we used it to understand Pandora listeners! We never said that their attitudes and perceptions — what they like or dislike about about Pandora versus FM/AM radio — was projectable to the entire population. That would have been absurd. That wasn’t our objective.

Every research methododology has its pros and cons, strengths and weaknesses. Understanding them helps us use the most appropriate methodology to achieve any particular objective. Even then, it’s never perfect.

Do I think that online delivery of “radio” programming will increasingly supplant transmitters and towers? Yes. Do I think it’s all going to happen in 5 years???  No way!