Counting print subscribers is easy. It’s time to bring that same precision to digital.

Photo (cc) 2015 by Dan Kennedy

By Dan Kennedy

In reporting on the newspaper business, there are few matters more obscure or maddening than determining paid digital circulation. My example for this morning is the Burlington Free Press, a Gannett-owned daily that, I wrote recently, will soon be printed in either Auburn, Massachusetts, or Providence, hours away from its home base in northern Vermont.

The change is the result of the giant newspaper chain’s decision to shut its printing plant in Portsmouth, New Hampshire, which will affect several other papers as well. Gannett’s standard defense of moves like this is that they’re shifting to digital, so print doesn’t matter that much. But as I observed at the time, the Free Press sells more than twice as many print papers as it does digital subscriptions.

My item prompted someone to send me a “confidential and internal” Gannett circulation report.

I’ll get to that. But first, let me explain what you’ll find if you look at reports filed with the Alliance for Audited Media, the organization to which newspaper publishers report circulation figures and which advertisers use to determine whether they’re getting a fair deal or not.

In order to make an apples-to-apples comparison, I’m going to use the Free Press’ AAM figures for the six-month period ending March 31, 2021, because that’s the last time they provided two different types of digital numbers. According to that report, the Free Press’ circulation was as follows:

  • Print: An average of 7,996 on weekdays; 10,295 on Sundays
  • Digital replica: 2,823 on weekdays; 1,407 on Sundays
  • Digital nonreplica: 1,407 on weekdays; 1,161 on Sundays

If you add the two types of digital together, which I’m told is a valid step to take, you end up with paid weekday circulation of 4,230, or just a little more than half of print circulation. Sunday digital is 2,568 — barely a quarter of Sunday print.

Now, what are these two types of digital? I am told by a highly knowledgeable source that they are what you’d imagine they are. Replica is an exact image of the paper, often called the e-paper. Nonreplica is the website and similar products, such mobile apps. But there is a huge problem with making that distinction: No newspaper that I’m aware of — certainly not the Burlington Free Press — charges separately for replica and nonreplica access.

The Free Press’ digital subscription offer

For instance, if you sign up for “unlimited digital access” to the Free Press you get, among other things, “full access to subscriber-only content on your desktop, tablet and mobile device” and “the eNewspaper, a digital replica of the print edition.” There is no separate sign-up for the e-paper; instead, everything is included in a digital subscription.

And lest you think this is a Gannett thing, the independently owned Boston Globe also reports separate numbers for replica and nonreplica. If you’re a Globe subscriber, as I am, then you know that a digital subscription gives you access to the website, which already includes the e-paper. There’s also a separate e-paper app available for mobile. So with the Globe, as well as with Gannett, it’s a mystery as to how they can claim to have separate paid digital numbers for replica and nonreplica, or for weekdays versus Sundays. And I do mean a mystery: I have asked a number of people over the past few years, including the AAM, and I’ve been unable to get a clear answer.

Now to get to that internal Gannett report: my informant wanted me to know that I was under-counting the Free Press’ paid digital circulation by quite a bit. And sure enough, the internal report showed that the Free Press’ paid digital in 2021 was 8,524 — double the combined replica-plus-nonreplica weekday number it had reported to the AAM and more than its weekday print circulation. By mid-2022 that had risen to 8,902. Unlike the AAM report, the internal figures show no separate breakdown for replica and nonreplica and no separate numbers for Sunday. Which makes sense, because that’s not the way digital subscriptions are sold. As I said, one price gets you everything — website, apps and e-paper, weekdays and Sundays.

Now, weirdly enough, the Globe’s reports to the AAM actually show higher digital circulation than what it has publicly announced. For instance, in February 2022, then-editor Brian McGrory told his staff that the digital-only paid circulation had reached 235,000 (up to 240,000 as of last fall). Yet according to AAM figures, the Globe’s paid weekday digital for the six-month period ending March 31, 2022, averaged 262,576, of which about 12,000 was replica. The Sunday digital figures were about 15,000 higher overall; again, about 12,000 was replica.

All of this is a shame because, before the rise of paid digital circulation, the AAM — formerly known as the Audit Bureau of Circulations — was the gold standard when it came to counting print. Digital has been a different story. Back in 2013, I reported that the AAM was double- and even triple-counting digital subscriptions depending on how many devices a particular subscriber was using, which was just nuts. (I also see that I quoted “Globe spokeswoman Ellen Clegg,” later the paper’s editorial page editor and now my research partner, co-author and podcast cohost.) I don’t think the AAM reports are that far off these days. But the Burlington Free Press and The Boston Globe show that whatever problems the AAM was having 10 years ago have not been fully resolved.

Paid digital subscriptions are now the standard method by which most newspapers measure their circulation. Print is not an afterthought, but it’s a lot less important than it used to be, and it continues to decline. What’s needed is for the newspaper business and the AAM to work together and come up with a standard method for measuring paid digital.

Correction: An earlier version of this post stated that the Free Press would be printed in either Worcester or Providence. In fact, the Gannett plant in question is located in Auburn, a suburb of Worcester.

Author: Dan Kennedy

I am a professor of journalism at Northeastern University specializing in the future of local journalism at My blog, Media Nation, is online at

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: