A R T I C L E S ~ A B O U T ~ O N L I N E ~ M A R K E T I N G
How many people are on the Internet and what are they like? Those two
questions are a major source of controversy right now. It seems that
everyone wants to know who's using the Internet and what they're using it
for. Madison Avenue is especially concerned with the Information
Superhighway; they can't sell ads unless they come up with some numbers for
the size of the audience.
The problem with calculating Internet demographics is that the net is
decentralized. No one controls access to the net, so it's very hard to
estimate the number of users. In this article, we'll examine the three
most-quoted surveys on the Internet and the World Wide Web. The numbers are
rough estimates and they're often used inappropriately. But the trends they
show are useful for making marketing decisions.
The MIDS Survey
This survey was conducted by Matrix Information and Directory Services (http://www.mids.org/index.html) in
late 1994. It is the most reliable of all Internet surveys I've seen, in
part because it was sent to Internet Service Providers rather than people
on the net. MIDS received 1463 responses from organizations that represent
roughly 10% of the people on the Internet.
The MIDS survey came up with the following numbers: 27.5 million people can
exchange Internet e-mail. Roughly half of these people (13.5 million) have
access to the Internet, and half again (7.8 million) have direct Internet
accounts via SLIP/PPP. These are the most accurate numbers yet on the size
of the net.
One problem with the MIDS survey is that it tallies accounts rather than
people. As many as one-third of the people using the World Wide Web have
multiple Internet accounts, according to one directory service. Therefore,
the number of people on the Internet could be much lower than the number of
accounts reported by MIDS.
Other interesting figures from the MIDS survey involve growth and gender.
The Internet has been doubling in size every year for the last six years.
While this growth rate will eventually slow, the popularity of the World
Wide Web is attracting more people than ever to the Internet. MIDS shows a
dramatic narrowing of the Internet gender gap. The ratio of males to
females is often quoted as ten to one; the new MIDS survey pegs this at two
After looking at all the data I could find, I feel most comfortable with
the following numbers, as of the summer of 1995: 30 - 40 million people
have e-mail, about 15 - 20 million people are on the Internet, and about 5
million people have access to the World Wide Web.
Just last week I cruised more than 100 malls on the World Wide Web. Mall
owners sell web storefronts to people who want to do business on the net.
Almost without exception, the mall owners were quoting a potential audience
of 30-40 million people, even though only a fraction of those people can
access the web.
For those trying to do business over the Internet, one thing is clear: you
need to make product information available via e-mail. If you are relying
on the web to reach your audience, you are missing as much as 90% of the
market. E-mail isn't as sexy as the web, but it's still the most effective
sales tool in the Internet kit.
Two other recent surveys deal with World Wide Web users only. Both surveys
are flawed; they were posted on the web and relied on voluntary compliance.
I haven't seen any attempt to subject these surveys to statistical
modeling, so please take the numbers with a shaker of salt.
The HERMES survey was conducted by the University of Michigan Business
School in Spring of 1995 and got 13,000 responses. The people who responded
are well-educated (80% had some college) and surprisingly well off (more
than 50% had incomes over $50,000). A similar web survey was conducted in
Fall of 1994 by the Georgia Institute of Technology. Of the 18,000 people
who responded, nearly half were between 26 and 30 years old and 90% were
(UPDATE: Hermes & Georgia Tech merged and now produce one survey, the GVU Web Survey, at http://www.cc.gatech.edu/gvu/user_surveys/User_Survey_Home.html. Another excellent web site for learning about Internet demographics is CyberAtlas at http://www.cyberatlas.com/.)
I don't want to quote these surveys at length because the numbers are so
unreliable. But there are some interesting indicators for those trying to
sell books over the net. The good news is that people use the web mostly
for entertainment or research. They enjoy reading e-zines and looking at
cool web sites, and they specifically cruise the web looking for product
The bad news is they don't buy online. As a reason for using the Internet,
shopping has finished dead last in every survey I've seen. People prefer to
buy in person. They even like direct mail better than the net. Part of the
reluctance to buy is due to security. The preferred way to close an online
purchase is through a toll-free phone call.
With so many Internet vendors now offering "secure transactions," there's
more than fear behind the weak retail performance of the net. Part of the
problem is the culture on the Internet, where almost everything is free.
Another problem for retailers is that most people are logging onto the net
from work. They might not mind planning their vacations on company time,
but they feel uncomfortable closing transactions right there in the office.
There's some hope for publishers, though. The purchase price of books is
relatively low, it's easy to get out information about your titles, and
online bookstores are some of the more successful Internet merchants. I
recently heard that 14% of the people responding to a survey had bought
books online while nearly half of them had bought books offline after
hearing about them on the net. Those numbers are encouraging, even if the
methodology behind them is weak.
STEVE O'KEEFE is author of the books Publicity on the Internet (John Wiley & Sons, 1997), and The Complete Guide to Internet Publicity(John Wiley & Sons, 2002). You can reach him by e-mail at info@PatronSaintPR.com.