|
In the interest of your time and our space, we will
limit the discussion of terminology to a brief summary of the most
important terms. Many terms will contain links to a fuller
description at one of the above sites.
Audience Terms
The first audience-related term to be used was "hit". This
term is still used today but mostly incorrectly. A hit is a record in a common (Unix) log file. Many
people use this term to indicate a visit to a site. A visit is a single session on a
site, comprised of as many pages as the visitor views. In most cases, each element on a
page (text, each graphic, each ad) represents a "hit". As such, a
single visit will have many hits included. Since there is no
relationship to hits and audience, we won't be revisiting this term.
Visitors are sometimes characterized as unique visitors.
This is equivalent to the traditional media term "exclusive cume".
While the term visitors counts every visit to a site, the term
"unique visitors" counts each person only once and does not credit
repeat visits.
The standard for measurement of traffic to Web sites is defined
by two sources, Nielsen NetRatings and comScore's Media
Metrix. Both sources have a significant share of market in the
audience measurement arena. They provide measures of visitors,
unique visitors, demographics and other data. Keep in mind that
unlike other media, these ratings services can only measure site
characteristics and, as of this date, are unable to measure
advertising audience. This is key as the Internet is the only medium
where the content audience and the advertising audience are not of
the same magnitude. We accept that we lose some audience in TV
through attentiveness or leaving the room and that not all magazine
readers peruse every page. But at least the opportunity to see the
ad is there and the drop off is probably only 10-20%. A single ad on
MSN, unless there is a roadblock or page takeover specifically
engineered to preempt all other advertising, will only reach a
portion of the audience at any given time as other pages on the site
will be serving ads from other advertisers at the same time. Please
note that the use of the term "reach" by these companies refers to
site reach or "cume potential". This should NOT be confused with the
reach of your campaign as measured through traditional media. Only
the third party ad servers can do that (see below).
Individual site visitor counts can be verified by I/PRO, the standard in
site audits. Individual site demographics (like a subscriber study)
can be measured by Tacoda.
A pageview represents a visit to a single page in a
session. While many in the Internet marketplace equate the term
pageview to advertising, the relationship is not one to one. It is
possible to have a page without any advertising or to have a page
that is abandoned before an ad is served. As such, when a site talks
about pageview, they are describing their total advertising
inventory available, not their audience size.
As with all traditional media, an impression is ideally
characterized as an OTS, or opportunity to see the ad. Thus, similar
to other media, an impression is a single exposure of a unique
audience member (user) to an advertising unit. While the MRC, IAB,
AAAA's ARF and others have worked to "normalize" this definition,
for technical reasons having to do with ad serving, there remain
several points at which an impression can be served. While there are
many discussions of this on the Web, the curious reader is probably
best served by reading the IAB information on this topic. For all practical
purposes, most media buys assume no difference site to site, and in
fact, it is expected that any differences will be resolved in the
short term. As with other media, an impression may also be referred
to as an exposure. Some also call it an adview.
Terms related to advertising or post advertising
actions
Third party ad serving: Ads generally come from a different
server than content. The server the ad comes from needs to control
inventory and be able to pick out and serve the next appropriate ad,
based on site parameters. It might literally be the next ad in the
queue, but if registration or some other data point indicates
demographic or sociographic aspects of a visitor, the ad server can
also skip down through ads until it finds the most appropriate ad
for, say an 18-34 year old female or one that the advertiser has
targeted to those who have indicated an interest in travel.
Given the way that the Web and the Internet are constructed, the
ad server can actually be anywhere in the world. Thus was born the
concept of the third party ad server (3PAS). Companies like
DoubleClick, Atlas DMT, Bluestreak, MediaPlex, 24/7 Real Media and others are employed by both
publishers, advertisers and agencies to serve their specific
advertising. Instead of the ad being embedded on the page, a tag is
placed there by the site pointing to the server containing the
assets (ads) that are to be served. The advantages of third party ad
serving for the marketer are many, but include: a) the ability to
make certain that all creative units are deployed at the same time,
b) an assurance that all metrics (impression and post impression
related) are on the same basis, c) an ability to go to one source
and see up to date performance of a campaign at any time, and d) the
ability to change advertising, either in part or on a wholesale
basis, quickly.
As the 3PAS companies serve all of the impressions for an
advertiser campaign, they become the source for counting the total
audience for the campaign, not only on a gross impressions basis but
on a net or unique basis (reach). The issue here is that the 3PAS do
not measure demographics. This resulted in work done over the past
several years to unify 3PAS information with ratings service
demographic data in order to provide accurate reach and frequency
estimates comparable to traditional media and which can be combined
with traditional media for purposes.
Reach, the number of users exposed to a specified campaign or ad
in a specified time period has been an elusive definition for the
reasons outlined above. However, there is considerable work being
done within the industry on this topic and credible first models
have been released by companies like IMS, Telmar and Atlas DMT. You can
also access R/F data through DoubleClick's Media Visor or from
NetRatings and comScore directly. While the R/F development today
only covers pre-buy or planning estimates, the tools are adequate
for estimating the reach and frequency of a campaign with a variety
of demographic possibilities. Within Telmar and IMS, this data can
also be combined with other media R/F data for purposes.
Clickthrough is a commonly referred to metric. It
represents the action of a user when they click on an ad or link and
are transferred from the site they are currently viewing to a new
location, usually the advertiser's site. More and more today, the
branding effect is taking place and those exposed to advertising are
coming to sites after the exposure has taken place. This is referred
to as post impression or viewthrough. This can take
place minutes, hours, days or weeks after the initial ad impression.
Through the use of cookies, a track record kept on an individual
user's computer recording items like passwords and visits to sites,
advertisers can determine if a visitor to a site has been exposed to
an ad previously, enabling credit of a view through. Post
advertising activity for many campaigns is split 50/50 between click
through and view through today.
The result of back end tracking is that evaluation of Interactive
campaigns can go far beyond the traditional CPM, or cost per
thousand impressions metric. We can of course measure
cost-per-click, or CPC. This metric however has become outmoded with
more sophisticated metrics made possible through use of the 3PAS
companies who track both clickthroughs and viewthroughs. The result
is that an advertiser can now measure CPR: Cost per response, CPA:
Cost per Acquisition (customer), CPS: Cost per sale, CPL: Cost Per
Lead, CPO: Cost per Order, CPT: Cost Per Transaction or what has
been called CPW: Cost per whatever (i.e., any of the above and
more). If you can devote a page to and action on your Web site, it
can be tracked.
The above data has put significant pressure on the Web to produce
meaningful ROI (Return on Investment). Three factors enter the
calculation. The cost of the media and advertising creative that
contributed the sales in question is the denominator. The dollar
value of those sales, minus costs, over the time interval assumed
for the advertising's effects on sales (short, medium or long term)
is the numerator. This calculation produces the dollar contribution
to earnings for each media dollar spent.
In addition to the DR aspects of the Web, significant evidence of
branding capabilities have been defined. There are many examples
within MSN Advertising and other sites of branding
efficacy on the Web. One of the great aspects of the Web is that, in
addition to being a media vehicle, it can be a feedback loop.
Companies like Dynamic Logic and Millward Brown
Intelliquest have the capability to do the type of attitude,
usage and awareness research on an almost instantaneous basis that
historically has taken months to field and months more to collate
results. Costs for this type of research on the Web can also be
significantly lower than traditional AU&A research.
In future articles, we will discuss concepts like the allocation
of Interactive budgets within the overall mix, planning vs. buying,
tracking and optimization and other issues relative to .
|