eAnalytics, the strategic analysis arm of Engage Technologies (NASDAQ: ENGA)
has issued its annual review of online advertising. Engage is a subsidiary of
CMGI (NASDAQ: CMGI). The report finds a modest decline in CPM rates, the basic
measure of the cost of advertising. The CPM rate is the price a website charges
an advertiser for showing one thousand impressions of an ad.
average CPM rate decreased from $35.13 in December 1998 to $33.75 in December
1999, about one percent each quarter. This is a smaller decrease than over the
previous twelve-month period, which saw rates drop six percent. Since AdKnowledge
started collecting the statistics in 1997 the rate has dropped a total of nine
percent. AdKnowledge concludes that rates are slowly leveling off.
and technology sites, which commanded the highest average CPM rates in 1998,
dropped to third place behind professional publications and corporate sites.
Sites covering money and finance, news, and travel have the next highest rates,
but also showed declines. Only shareware sites (which had the lowest rates in
1998) and shopping/transaction sites showed increases.
dropping rates represent a loss of revenue for both ad agencies and website
publishers, the former, at least, are feeling no pain. That is because the number
of web sites that court advertising has risen 135 percent since the start of
1999, and more than six hundred percent over the two-year period.
publishers may feel some pinch with the drop in rates, but it has been conventional
wisdom for some time that ad revenue alone will not sustain most sites. The
increase in the number of sites seeking advertising bodes well for the ad agencies,
but ultimately their revenue derives from placed ads, not opportunities to place
them. As inventory, the number of available places to put ads, goes up rates
are bound to decline. Therefore, the relatively small decline in rates looks
to us to be excellent news for the advertising industry.
continue to believe that Internet advertising is a growth industry for both
publishers and advertising agencies, albeit a more modest one than in the frontier
days of the Internet.
the future, we see at least three important forces at play:
- CPM rates and inventory -- In a steady-state situation, CPM rates and inventory
would look stable over a long enough time span, but might tend to fluctuate
over shorter times, much as do predator-prey populations in the wilderness.
(First the wolves kill off almost all the deer, but then with no venison to
eat the wolf populations decline, giving the deer a chance to replenish. Then
the wolves have easy pickings, so their numbers rise, and the cycle repeats.
This is a classic example in college courses on differential equations. We
would not expect such dramatic cyclicity to appear in Internet advertising,
however.) Whether such a steady state can be reached is affected by the other
- The general state of the economy -- The second factor influencing advertising
growth is the general state of the economy as reflected in the interest of
companies to advertise.
- The technological trend toward more precise ad delivery -- Advertisers
continue to roll out means for assuring advertisers that ads will be delivered
only to surfers who are likely to respond to them. The effect of such technologies
is to allow advertisers to reach more of the people they want to reach with
fewer overall impressions. As long as advertisers are willing to pay more
for this on a CPM basis, and since a given advertiser thus uses fewer overall
impressions, available inventory will rise. And, as noted above, a rise in
inventory will lead to a decrease in CPM rates.
there is a prediction to be made here it is that within a few years some of
the basic parameters of these processes will be determined experimentally, and
students of differential equations will be working homework problems based on
A company planning to build a business based on revenues from advertising on
a website should develop its business plan based on slowly decreasing CPM rates.
The same 10 million monthly ad impressions - not an unreasonable number for
a small but aggressive website -- would have brought in $13,000 less in December
1999 than in the previous December.