The Mysteries of Google
by Dyanna S. Culp (Temos)
Google is hands down the #1 United States search engine of choice.
Why? They offer a massive database, spilt second search response time
and relevant results. This is great if you're searching the Web, but
their dominance has turned many webmasters into neurotic Google position
watchers.
If you don't rank well on Google you might as well not exist, especially
now that they provide data feeds to AOL and Yahoo. Depending on the
source, it's estimated that Google now covers 60- 75% of all Web searches
done in America. Keeping up with, and understanding, the mysteries
of Google is critical for a Web site's success.
GOOGLE BEGINNINGS
Google hit the Web in 1998. In 1998, two Stanford computer science
students (Sergey Brin and Larry Page) began identifying Web link structure
patterns for a class project. Their analysis of "backlinks" (pages
that link back to a site) inspired the basis for the first Google
algorithm ranking system.
Google is a play on the word "googol" first used by the mathematician
Milton Sirotta. It represents the number one, followed by 100 zeros.
The term Google emphasizes the mission to collect and organize vast
amounts of online data. When the two Stanford students presented their
search engine idea to their first investor, they received a check
made out to "Google, Inc." They rushed out to formally set up the
company name, and a bank account, so they could cash that first check.
Google was a late starter, but in just five years they came to dominate
the Web in America and are extending their reach worldwide. Google
provides search services (or data feeds) to over 130 search engines,
portals and corporate Web sites in more than thirty countries. You
are seeing Google search results if you visit AOL, Cisco, Netscape,
Yahoo, Sony and hundreds of other sites. The first three Google years
were heaven for searchers and for webmasters who took a little time
to study Google for ranking strategies. The fourth year, as Google
came to dominate the search scene, it held the success or failure
of any online business in its grasp.
During the year 2003 webmasters experienced nervous breakdowns as
the Google System underwent a massive change. Google is radically
altering the brains behind their system's ranking algorithm in an
attempt to offer "more relevant results" and to stymie webmasters
who've figured out how to rise to the top of the ranks after a Google
Dance. Google results were a mess for most of spring 2003. Sites online
for over a decade disappeared completely. Sites that were nothing
more than link farms sometimes rose to the top. The infamous "relevant
results" were gone. Google quickly lost an estimated 10 % of their
search audience. This summer things settled down (improved), but Google
still has some strange quirks in their results, especially after a
Google Dance.
THE GOOGLE DANCE
What is the Google Dance? Google crawls about eight billon Web documents
(including HTML, PDF, RTF, Word, Excel, PostScript and PowerPoint
files) to update its search indices. This data is collected at eight
primary data centers and then distributed to thousands of computers
in many locations. The Google Dance occurs when new information is
being distributed and the eight data centers do not have the same
final results. During the dance, dead URLs are dropped, titles and
descriptions updated, and ranking positions changed. It might seem
reasonable to think that your personal PC is always fed by the same
center, but that is not the case. So one hour you might see your optimized
site at #3 for "heirloom roses," and the next hour (to your dismay)
you might be at #45.
You can check your site keyword status at each of these eight data
centers. If the results vary radically at the different urls - then
you're experiencing a Google Dance.
www-sj.google.com
www-fi.google.com
www-cw.google.com
www-dc.google.com
www-va.google.com
www-ab.google.com
www-ex.google.com
www-in.google.com
Until spring of 2003 the Dance predictably occurred roughly once a
month. After two or three days the dance stopped, and ranking results
were pretty stable until the next month. For about three weeks each
month webmasters could breathe easy - they knew where they stood.
Well, things have changed, and a new acronym is now used to describe
many webmasters "GDS" or Google Distress Syndrome. Google is now updating
its index more often and without any discernable time pattern. Many
speculate that Google is moving to a constant spider/ update process
where large established sites are visited almost daily. This appeared
to be the case for about a month, but now the spidering is becoming
more erratic. Now why would this concern anyone?
The problem is not the Dance time frame, but the quirkiness of the
new Google ranking algorithm. Google is implementing new spam filters
to penalize sites that do not adhere to the Google webmaster rules.
They are also experimenting with new ranking methods for analyzing
backlinks- the number and type of links that go to your site. These
changes might be welcomed if they were working properly. But the new
system has problems, and many good Web sites are being penalized.
LINKING AFFECTS ON THE NEW ALGORITHM
We've known for a long time that Google ranking was influenced by
incoming links and that exchanging links with a low ranking site,
or link farm, could radically reduce a site's ranking. But that influence
is increasing with the new system. As Google activates new changes
in how links affect their overall algorithm, those with highly optimized
Web sites may experience severe GDS. If the links aren't there, an
optimized site will find itself in a much lower position than it enjoyed
six months ago.
Page Rank Algorithm matches a site's relevance to its incoming links.
It is the heart of Google's ranking system. Google was the first,
but now most major search engines use link popularity as part of their
ranking system. If you have 10 sites linking to you which Google ranks
as important, relevant, popular sites, this carries more weight than
a hundred links from Web peons or link farms. Those links won't hurt
you, but they won't do you much good. However, a link on your site
to a low ranker could actually drop your site's position. So, be careful
with whom you exchange links.
The descriptive text in incoming links is now critical to ranking
position. Even the hyperlinks within your own site should be carefully
worded. Let's say your company name is Taylor's Nursery and you want
to rank high for your heirloom roses. You have other sites linking
to you, and their hyperlinks are followed by descriptions of your
company or Web site. Having the keywords in the description is not
enough. They must be in the hyperlinked text, and the closer to the
front of the text, the better.
Poor text link choice: Taylor's Nurseries
Better choice: Taylor's Nursery Heirloom Roses
Best choice: Heirloom Roses Taylor's Nursery
Google software engineer Matt Cutts: "Google's PageRank search technology
works by first identifying the link structure of the entire Web, then
ranking individual pages based on the number and importance of pages
linked to them."
PART TWO: Content
Affects on Google Algorithm
|
Printer
Friendly Version