Digital circulation figures are an absolute mess
We've written quite a bit at Poynter about how newspaper circulation numbers are basically meaningless now. The Alliance for Audited Media tries to provide a helpful framework for reporting digital readership, but the ways we consume news are so varied that it's tough to nail down exactly what should count.
AAM acknowledges as much, cautioning against reading too much into overall circulation figures, particularly when it comes to generating top 10 lists and such (the organization itself stopped publishing a top 25 list last year). But that doesn't keep newspapers from celebrating misleading numbers to whatever extent they can, so it didn't stop me from trying to figure out if the numbers obscure some troubling trends.
(As he does every six months, Andrew Beaujon tackled AAM's report when it was released last week, pointing out how branded editions allowed the New York Times, USA Today and Orange County Register to hyperinflate their numbers.)
I decided to zero in on AAM's "digital nonreplica" category because it reflects mobile app use and paywall subscribers, two of the major growth areas for newspapers trying to counter print declines. Performance in that category could be a good indicator of a newspaper's digital success. (A separate "digital replica" category counts mostly PDF replica use.)
But even comparing newspapers by digital circulation alone can be misleading. When you do compare digital counts across newspaper, you end up with weird graphics like this one that don't reflect actual readership or audience reach. That's because only apps, paywalled websites, PDF print replicas and e-reader editions count in circulation figures (visits to free websites don't). Moreover, newspapers can pick and choose which digital categories to report.
So digital circulation varies wildly from paper to paper due to differences in what they choose to report and differences in the types of products they offer. Further complicating things is the fact that readers can be double-, triple-, or even quadruple-counted if they're print subscribers and use various apps and digital products, too.
Rather than ranking anyone, then, let's look at how average Monday-Friday digital nonreplica circulation has changed for individual papers in the last six months and 12 months:
USA Today's "digital nonreplica" circulation was 1,484,078 in last September's report. In last week's report, the number had dropped 8 percent, to 1,365,388. The category, which mostly reflects app use, actually contributed to an overall decline in reported circulation. (Without its new branded "butterfly" editions, USA Today's total circulation fell since September's report by almost 10 percent, from 2,862,229 to 2,587,103.)
But USA Today's story celebrated the circulation figures:
Time spent on USA TODAY's mobile and tablet apps also continues to contribute to a circulation increase. "We have more people who go to our apps and spend more time on (them)," Kramer says.
A USA Today spokesperson told me via email that it's "apples to oranges" to compare March's numbers with September's. But year-over-year comparisons make even less sense because app use wasn't counted by USA Today in March 2013, when AAM's bylaws still required 70 percent of circulation to be paid (USA Today's apps are free). The result: USA Today's huge, misleading year-over-year circulation jump that I addressed in October.
So what could account for the drop in the last six months that some major newspapers seemed to observe?
I asked Neal Lulofs, AAM's executive vice president for marketing and strategy, if it was a fair comparison. He said it made sense to compare September's reporting period with the current one given changes that weren't yet in effect a year ago, but noted that each period sees some natural fluctuation as newspapers adjust to AAM's complicated guidelines.
"Our rules in this area are pretty dense," Lulofs wrote via email. "We think some publishers are still adjusting to them, so there may be some self-correction of the numbers period to period."
Moreover, Lulofs said the AAM board is considering changing how it counts digital circulation. A reader currently counts as a circulation unit as long as he or she accesses an app or paywalled website at least once per month. If AAM changes that, the reader would count only for the days he or she accesses it — more like how print circulation is calculated.
Lulofs speculated some newspapers might already be reporting digital circulation based on that standard in anticipation of a change that would likely decrease the number of readers they're allowed to report as circulation.
(Among other potential reasons for the fluctuation: ups and downs from promotional subscription rates and the introduction and retirement of various digital products.)
Alternatively, these numbers really could mean digital circulation has plateaued for some newspapers. But it's hard to tell, especially because newspapers continue to write press releases disguised as news stories whenever the numbers are released rather than providing meaningful context.
Of course, AAM primarily serves advertisers, who are increasingly looking for granular data about audiences. That's why metrics like total circulation and even total digital circulation don't make much sense anymore. As John P. Murray, Newspaper Association of America vice president of audience development, wrote in November:
For newspapers, simple is no longer better. Timely, comprehensive data on each platform is better than attempting to translate everything into a single metric such as total circulation. Now is the time to make sure we are presenting the right data to demonstrate the opportunities your newspaper has to offer your customers.
So here's the moral of the story: Digital circulation numbers might mean something when it comes to taking the newspaper industry's temperature, but it's exceedingly difficult to find out if they do given the pace of change, both in the products newspapers offer and how AAM counts readers. It's almost impossible to find apples-to-apples comparisons in such a bizarre fruit salad.
Related: USA Today’s circulation up 67 percent? Newspaper industry makes comparisons increasingly difficult (Oct. 2013) | Why USA Today’s huge growth in digital circulation isn’t what it seems (Oct. 2013) | Daily newspaper circulation totals ‘do not capture the full story’ anymore (April 2013)