Three surveys, three total figures. Why the numbers don't add up
Pick a number between 1 and 25 million.
Congratulations, you now have an estimate of how many people are on the Internet--and a guess that might be as close as pricey market research surveys.
Market researchers cannot agree on the number of Net surfers, creating confusion for marketing and media executives trying to map Internet strategies.
Who's right? All three.
TROUBLE WITH DEFINITIONS
"How the user is defined is really going to define what numbers you [get]," said Daniel Campbell, research director at Find/SVP, which published its Net survey in January. "It's not an issue of good or bad surveys. It's an issue of how do you define 'access'?"
Size is important, given that marketers and media must make business decisions based in part on just how mass the new medium is--and what to do with the more than 30,000 business Web sites. (That number, too, is in dispute, but that's another story.)
O'Reilly, working with market researcher Trish Information Services, and Find/SVP honed in on core Net users, leading to lower numbers.
In a study prepared for CommerceNet, a consortium of technology companies and others promoting electronic commerce, Nielsen defined the potential market more broadly. It found 37 million people in the U.S. and Canada who claimed "access" to the Net--including those, for example, who never used it but had friends who might give them access.
Extrapolating, Nielsen came up with 24 million "users"--a figure that Paul Lindstrom, VP at Nielsen Media Research, still says probably is too conservative.
Those who have worked with the Net for some time have come to expect such inconsistencies.
"It would be nice if there was some consensus," said Dick Hackenberg, interactive account director at TBWA Chiat/Day, Venice Calif.
Still, it's never good when a leading survey takes a hit from one of its researchers, which happened to Nielsen.
Donna Hoffman, a marketing professor at Vanderbilt University who initially consulted with Nielsen on the study, contends Nielsen exaggerated its numbers by, among other matters, overemphasizing the number of wealthier and more educated consumers.
Mr. Lindstrom stands by the methodology, and Nielsen has responded to Ms. Hoffman's criticisms. At the same time, however, the company has held up release of another Internet survey while it reviews the data.
Just six months ago, marketers and media had little hard data on the size of the Net. Now there may be too much. No company, Nielsen included, is yet the "Nielsen" of new media.
STRADDLING THE FENCE
Even the surveys' backers are straddling the fence: America Online supported all three.
"Consumers [of research data] need to question the research companies," said Find/SVP's Mr. Campbell.
But comparisons aren't easy because neither the producers nor buyers of research have decided what criteria are most important.
Nielsen, for example, is leaving it to the market to decide whether to size the Net based on the number of people who went there over the past three months (24 million) or over the past 24 hours (7 million).
So how big is the Internet?
"Don't ask me, because I don't know," said Ms. Hoffman, who probably has scrutinized the data as much as anyone.
Debra Aho Williamson contributed to this story.
Copyright February 1996 Crain Communications Inc.