NEW YORK: New research from the Audit Bureau of Circulations and Newspaper Services of America calls into question the effectiveness of a national reader per copy number (RPC) -- a figure supplied by the Newspaper Association of America (NAA).
The report, which was presented last week at the Worldwide Readership Research Symposium in Prague, suggests that some newspaper salespeople (if not publishers) use simple general formulas to arrive at total readership estimates for specific newspapers. This method overrides available local estimates or fills in the gaps for missing data, said the report. Furthermore, it underestimates the value of the medium.
"We're not disputing the accuracy on a national basis," said Kevin Campbell, director of marketing and sales at ABC. "But the problem is, newspapers are local market mediums."
The report objects to taking the national average for readers-per-copy, which is approximately 2.3 daily and 2.5 Sunday, to calculate specific readership stats. For instance, if Paper X's paid daily circulation is 100,000, then 100,000 might be multiplied by 2.3 to arrive at 233,800 readers.
John Kimball, senior vice president and CMO of NAA, contests that the figure is used for that purpose. "There isn't a newspaper I know of that would do that," he said. The RPC, he said, "is not a static number. That's why [newspapers] do individual market research."
ABC points to several pages on the NAA site (here and here) where the NAA uses RPC data and the modeling approach.
But it's because of that market research -- with over 30 suppliers operating in the industry -- that some newspapers fall back to the formula, Campbell said. "At the same time, advertiser access to syndicated and custom readership can be spotty, slow, and inconsistently calculated or reported when finally available," said the report.
The white paper advocates a different method to arrive at readership numbers, making it "newspaper specific."
For their model, ABC and NSA reviewed 237 Reader Profile reports (from ABC) using the following formula: newspaper specific readership estimate divided by paid circulation equals newspaper specific RPC.
The study found a wide range in individual newspapers from 1.8 to 4.4 RPC. From this base, 38 daily newspapers confirm around 2.3 RPC -- the national average -- but 46 papers report a lower RPC and 153 papers show a higher RPC than the national average.
In other words, using 2.3 RPC to arrive at readership is only accurate 16% of the time. In the majority of these cases (65%), using the national RPC "underestimates daily individual newspaper readership by as much as 91%," according to the report.
When the team applied their model to Sunday, the range was even greater. A review of 217 papers reporting Sunday readership shows a range between 1.8 to 7.1 RPC.
Throwing out 7.1 RPC -- it was eliminated from the sample because it may be an anomaly -- the report finds that 65% of Sunday newspapers are below the national average of 2.6 RPC.
So what does this all mean? The study looked at one national company in the telecommunications industry that is a leading newspaper advertiser. Excluding insert advertising and those papers that do not have a Reader Profile report, this company spent $107 million in newspapers.
Using specific local information, the study found that total daily readership for that advertiser was 58,230,800 adults -- or 9.1% higher than the number found by using national RPC. Using the national RPC number also yielded a higher CPM.
"In the age of increasing accountability and accuracy, the authors believe financially responsible professionals can not make sound investments using modeled readership estimates," the report concluded.
Jennifer Saba (jsaba@editorandpublisher.com) is associate editor at E&P.
