Getting quality feedback from readers is too important to relegate to just an occasional quest.

By William Dunkerley

A recent anecdotal survey suggests that many editors actively survey reader interests every one to three years. That's probably fairly reflective of practices in general -- except, of course, those who either don't survey at all or hardly ever do. They didn't opt in to the study.

I'd like to suggest that getting more frequent reader feedback can be beneficial.

The frequency of gathering metrics should reflect how much variance may be occurring. So if you are going to measure the height of the Eiffel Tower, for instance, there's not much need for an ongoing program. If you're driving a car, however, it's a good idea to check your speedometer often, lest you get caught in a speed trap going too fast.

How Often Should You Survey Readers?

It's a pretty sure bet that there are subtleties going on with your readership that are left undetected by gathering metrics at one- to three-year intervals.

My recommendation is to get some sort of quantifiable feedback with every issue. That doesn't mean doing a long, comprehensive survey each time. There are other ways to keep an objective hand on the pulse of reader interests.

Here's a case example from a publication I worked with a while back. The editor wanted to know how readers felt about all the individual articles she was publishing. Her hobby magazine had a circulation of over 150,000. The questionnaire I developed listed all the articles in an issue, and asked the recipient to rate each one on a scale of 1 to 5, with 1 being the best rating.

I also included the advertiser index on the list because it changed very little from issue to issue, and there was no reason for readers to rate it differently month to month. Therefore, it served as an indicator of sample error.

I started by sending the questionnaire to a list of 300 readers randomly drawn from the 150,000. The response rate turned out to be about one-third every month. I found that there was almost no random variance in how readers rated the ad index from month to month. That told me that my sample size was probably large enough for this relatively simple survey. If there had been greater variance on the ad index question, I would have evaluated whether that much sample error was acceptable when viewing answers to the other questionnaire items. If it was not, I would have increased the sample size until I reached an acceptable level.

What Are the Benefits of Frequent Surveying?

This survey program gave the editor the answers she sought regarding the articles. But there is a lot more insight that can be mined from the results, examined over time.

The first point is the response rate. In the above example it was 33 percent. But I've seen other publications that could pull only five percent. All things being equal, that says something about the overall interest of readers in your publication.

You can also analyze the relative interests in various categories of articles. Actually, that is something that would be handy to include in your large, comprehensive survey -- i.e., "Please rate your level of interest in the following topics that we cover...." Armed with that as background, you can go to work with the results from your perpetual surveys.

To do this kind of analysis, start with perhaps a year's accumulation of data. For instance, if you publish a news magazine, you might track the ratings of articles about politics, crime, sports, travel, fashion, global affairs, etc. Aggregate the ratings readers give articles in each category. Now compare that with the areas of interest found from your comprehensive survey. The big survey tells you where their interests lie; the perpetual survey tells you how you're doing in serving their interests.

Other Modes of Analyzing Results

Other than subject categories, there are other useful schemes for aggregating results. Compare liberally illustrated articles with ones that are text heavy. Compare position within the publication. Compare various writing styles. You likely can add more to this list.

Are you in tune with whatever seasonality there is in reader interests? Most editors are very perceptive to the obvious trends. But there may be subtler periodic swings that go undetected by just commonsense observation. The perpetual survey will perhaps open your eyes to something unexpected.

All this will give you valuable editorial insights that may not be readily apparent otherwise. And if you make this survey a truly perpetual practice, you'll be able to track trends that emerge over time. That will give you an opportunity to fine-tune your coverage, and even to forecast the trajectory of changes in reader interests. That will be increasingly important as we face growing content competition from non-publication Internet sources.

William Dunkerley is principal of William Dunkerley Publishing Consultants, www.publishinghelp.com.

Add your comment.