The Numbers Game

As they do every six months, last week the Audit Bureau of Circulations (ABC) reported newspaper circulation averages for the "FAS-FAX period" from April through September. The news didn't look good, with weekday numbers down 4.64 percent year-over-year, and Sunday down 4.64 percent. This represented an acceleration of the prior-year pace of decline, which was 2.6 percent weekdays and 3.5 percent Sundays.

As usual, the industry tried to put a good spin on the numbers, as summed up by Russell Adams in the Wall Street Journal:

But the reality is in some ways less bleak than the latest numbers indicate: Some newspapers have raised newsstand prices, curtailed discounted copies and halted delivery to the least profitable customers. Also, while print circulation has been declining for years as readers continue their mass migration to the Web, many publishers point out they are reaching more readers than before through print and online. The problem for publishers is the printed paper commands higher ad rates than the Web so even as more people read newspaper content, the papers pull in less money.

Some newspapers were able to point to gains in overall combined audience of online and print readers. In particular, Matt Baldwin, V.P. of Research at MediaNews Group, wrote an indignant memo claiming growth in the total audience of the Denver Post (owned by MediaNews) and Rocky Mountain News (managed by MediaNews as part of a joint operating agreement). This came after the Post as well as the Rocky had each reported their print circulation drops the other day without, apparently, getting the online audience spin from their V.P. of Research.

An obvious question about this combined audience approach might be, aren't we comparing apples and oranges? The print portion of the averages reported (in the first link in the previous paragraph) are seven-day readers. A one-day reader counts as only one-seventh of a reader in that average. But the are adding to that the seven-day "unique visitor" number, in which a one-time, one-day web site visitor counts as one full reader. Apples and apples would demand adding only the one-day unique visitor average, which is undoubtedly much lower. Moreover, as acknowledged by Baldwin, some print readers are also site visitors, and nothing has been done to eliminate the overlap.

A set of web audience metrics that looks more reliable comes from the National Newspaper Association (NAA), which reports:

Newspaper Web sites attracted more than 68.3 million unique visitors on average (41.4 percent of all Internet users) in the third quarter of 2008, a record number that reflects a 15.8 percent increase over the same period a year ago, according to a custom analysis provided by Nielsen Online for the Newspaper Association of America.

The NAA recognizes that both the election and the economy are contributing toward this rise, but their stats show pretty similar reach during July, August and September, which is not what you'd expect if the trend were due mainly to current events. In the NAA's findings, the average unique visitor made eight or nine visits, however, so this is not seven-day readership. In fact, if you examine their stats closely, the NAA is saying that 41 percent of web users spend about 45 minutes a month, or an average of only 90 seconds per day, at newspaper sites. If you look at it that way, it's not much of an audience. In fact, the top half-dozen or so domain names all individually outpull the entire newspaper industry in unique visitors.

Still, there seem to be some optimists out there. Neither the revenue slide nor the circulation decline seem to bother Richard Siklos at Fortune in recommending a bet on printed newspapers.