Quality Leads and Search Engines - Part Three
At the present time, there really are no
search engines, per se, capable of
assessing the quality or value of any of
the Web Sites they engage and analyze.
This would call for a level of artificial
intelligence which is beyond the
performance limits of existing software
programs - even the new generation of
so-called "smart" search programs such
as Clever or Google.
In general, the algorithms employed by a
search engine to probe Web Sites tend to
focus on certain kinds of properties.
These properties usually involve one of
several possibilities - namely, they are
preoccupied with either formal features
of a Web Site, or with features which can
be quantified in some way, or some
combination of these two factors.
For example, an algorithm may be
fashioned to zero-in on the kind of meta
tags being used at the beginning of a
Web Page, along with whatever
information may be bracketed within
such tags. This would be an instance of
what is meant by a formal feature of a
Web Site since, for the most part, this
method is concerned solely with the way
in which a Page is coded or with the way
the description of the Page is worded and
not at all with what the quality of the
primary content of a Page may be.
A Web Page might have first rate HTML
coding, meta tags, title tags and so on,
and, yet, have little or nothing of value to
offer to a Web surfer. Therefore, the fact
that such a Page has a high ranking
within a given search engine does not
necessarily mean this Page will display a
high quality of information or content.
Alternatively, an algorithm might be
constructed to examine, for instance, the
number of other Web Sites which link to
a given Web Page. This is a quantifiable
feature and can be used to generate an
index of "popularity" of a given Page.
Although popularity is not necessarily
synonymous with quality, it might be
used by some to serve as a basis for
inferring that the reason why various
Web Sites are linking to the Page in
question is because people consider the
Site to be of value, interest and use to
them. On the other hand, an algorithm
that keys in on this quantifiable feature
(i.e.,the number of other Sites which link
to a Page) may only serve to lead us into
a circular argument.
More specifically, one reason why a
given Web Site may be popular is
because such a Page is one that keeps
coming up in a keyword search of data
bases employing this kind of an
algorithm. In other words, if a Web Page
has high placement in a given search
engine because a number of other Web
Sites link to that Page, and this is one of
the primary criteria used by a search
engine to rank Web pages, then no one
should be surprised that other Web
Pages - no matter how good they may be
- will not be visible in the placement
hierarchy of such a search engine or
directory as long as these latter Pages do
not get linked to by a large number of
other Web Sites.
This helps to establish a sort of
self-fulfilling prophecy. A Web Page
having high visibility in a search engine
due to the fact the Page possesses the
formal or quantitative features for which
a search engine is looking will continue
to enjoy high visibility in that search
engine precisely because it conforms to,
and reflects, the logic by which such a
search engine operates.
As a result, more and more Sites may
link to this Web Page simply because this
Page is being given visibility by the
search engine or directory to the relative
exclusion (i.e., visibility) of other Web
Pages and, in the process, a false picture
is being given of what is out there in
cyberspace. The way in which a search
engine creates visibility (i.e., appearing
in, say, the top twenty "matches" of a
keyword query) tends to place
constraints on choice, and in doing so,
the search engines and directories
generate an illusion of popularity when,
in fact, the popularity is largely contrived
through the limited choices which
visibility underwrites.
Web Sites are linking to the Page
because the owners of those Sites are not
being shown there may be other Web
Pages in cyberspace which may have a
lot more to offer than the particular Page
to which a given search engine is
directing them. As such, this Web Page is
"popular" by default since few other
alternatives are being given visibility in
the search engine simply because these
alternatives lack a quantifiable feature
which may have little to do with value,
quality or excellence of Web content.
A similar kind of problem arises in
conjunction with some of the
pay-for-keyword arrangements now
being used by a few search
engines/directories. In other words, if
someone has paid to have proprietary
control of one or more keywords, and a
surfer enters this keyword into a search
of the data base of a company operating
on this keyword-for-hire policy, then
what is the surfer getting when the
"search" result provides the URL of the
individual who is paying for the keyword?
The surfer is, in reality, getting a biased,
skewed, and, most importantly in the
present context, a somewhat circular
result. This is so because the keyword
search being used really has not
examined any of the other URLs in the
data base for true relevancy to the
interests of the surfer, but, instead, has
generated a result that is a function of an
artificial relevancy rooted in a
behind-the-scenes commercial
transaction restricting the "matches" of
the search engine to an URL for which a
client of the search engine has paid so
that the keyword in question will lead to
the client's URL.
The fact of the matter is that all search
engines produce biased results of one
sort or another even if no
keyword-for-hire policy is involved. The
bias in question in any given search
engine is a reflection of the assumptions,
values, and interests of the person or
persons who have constructed the
algorithm(s) to which a given spider or
search bot gives expression.
If the algorithms governing the operation
of a spider or bot program are keyed to
certain kinds of meta tags or header
statements, then this is what will be
reflected in the position rankings arising
from keyword queries of the data base
that has been built from the information
returned by such a program. If the
algorithms underlying a search engine's
activity have been constructed to give
weighted or preferential treatment to
Web Pages which are linked to by, for
whatever unknown reasons, lots of other
Sites, then URL placement in such a
search engine will conform to this
preference of the algorithm.
The order of appearance of URLs that
emerge in response to a keyword query
are not random selections from the data
base of a search engine. The placement
of URLs within a "search" return to a
keyword inquiry are a direct,
determinate result of, among other
things, the extent to which the HTML
coding for a given Web Page satisfies the
formal and quantitative conditions,
values and priorities that have been
worked into the program organizing the
search engine data base being queried.
This is the reason why various kinds of
search engine positioning software and
services are capable, within certain
limits, of helping clients to improve their
ranking within a given search engine or
directory, quite independently of having
to worry about improving the quality or
value or usefulness of a client's Web Site.
Search engines have almost nothing to
do with being able to identify the quality
or value of a given Web page, and almost
everything to do with stuffing Web Sites
into a set of pre-fabricated data-profiles
which reflect the properties of the
algorithm(s) being used to generate and
organize the data base underlying a
search engine.
The current situation of most search
engines and directories is somewhat like
the story of an obviously inebriated
neighbor who was observed, one night,
to, first, go down on hands and knees
near a light standard, crawl about for a
while as if looking for something, and,
then, get up, go to some other area near
the light standard and repeat the whole
process.
An individual who had been watching the
whole thing decided to see if he could
help the poor guy out. So, he walks over
to the drunk person and says: "You seem
to be looking for something, did you lose
your wallet or house keys?"
The inebriated man nods his head and
says one word: "Wallet."
The good Samaritan responds with:
"How do you know you lost your wallet
around here?"
The drunken neighbor wavers rather
unsteadily, looks at his neighbor for a
moment, and explains: "Well, actually, I
have no idea where I lost my wallet, but
there is some light here, and, so, this is
the only place where I can see what I'm
doing."
Search engines and directories are like
the drunk of this story. They look in the
places where their algorithms shed some
light, irrespective of whether what needs
to be discovered can be found within the
spaces illuminated by such light.
While search engines and directories do
"search" for information within their
respective data bases, these services
might be more properly referred to as
"link generators". A person stimulates
the search engine or directory to set its
algorithms into motion by entering one
or more words, and a black-box
operation chugs away to produce a
hierarchical set of links which may, or
may not, contain what the person needs.
The individual can take one, or more, of
these leads and proceed to check things
out. If such leads don't work out, the
person can either return to the search
engine/directory and try some other
keyword stimulus, or, if dissatisfied with
the previous results, perhaps, turn to
some other search engine to take at
spinning the latter's algorithmic roulette
wheel.
When search engines and directories
were, by and large, the only game in
town through which to locate Web Pages
that might be relevant to one's search,
one was forced to try to live with the
limitations and problems entailed by
search engines and directories. This no
longer needs to be the case.
| Part 1 | Part 2 | Next | Part 5 |
Return to Internet Menu
|
