The Ever-Elusive Algorithm
ne element of search marketing that has many people scratching their head in confusion is the
algorithms that actually determine what the rank of a page should be. These algorithms are pro-
prietary in nature, and so few people outside the search engine companies have seen them in their
entirety. Even if you were to see the algorithm, you’d have to be a math wizard to understand it. And
that’s what makes figuring out the whole concept of optimizing for search engines so difficult.
To put it as plainly as possible, the algorithm that a search engine uses establishes a baseline to
which all web pages are compared. The baseline varies from search engine to search engine. For
example, more than 200 factors are used to establish a baseline in the Google algorithm. And
though people have figured out some of the primary parts of the algorithm, there’s just no way to
know all of the parts, especially when you realize that Google makes about half a dozen changes to
that algorithm each week. Some of those changes are major, others are minor. But all make the algo-
rithm a dynamic force to be reckoned with.
Knowing that, when creating your web site (or updating it for SEO), you can keep a few design prin-
ciples in mind. And the most important of those principles is to design your web site for people, not
for search engines. So if you’re building a site about springtime vacations, you’ll want to include
information and links to help users plan their springtime vacations.
Then if a crawler examines your site and it contains links to airfare sites, festival sites, garden shows,
and other related sites, the crawler can follow these links, using the algorithm to determine if they are
related, and your site ranks higher than if all the links lead to completely unrelated sites. (If they do, that
tells the crawler you’ve set up a bogus link farm, and it will either rank your site very low or not at all.)
The magic number of how many links must be related and how many can be unrelated is just that, a
magic number. Presumably, however, if you design a web page about springtime vacations and it’s
legitimate, all the links from that page (or to the page) will be related in some way or another. The
exception might be advertisements, which are clearly marked as advertisements. Another exception
is if all your links are advertisements that lead to someplace unrelated to the topic (represented by
keywords) at hand. You probably wouldn’t want to have a site that only had links from advertise-
ments, though, because this would likely decrease your search engine ranking.
The same is true of keywords. Some search engines prefer that you use a higher keyword density
than others. For all search engines, content is important, but the factors that determine whether or
not the content helps or hurts your ranking differ from one search engine to another. And then there
are meta tags, which are also weighted differently by search engines.
So this mysterious baseline that we’re talking about will vary from search engine to search engine.
Some search engines look more closely at links than others do, some look at keywords and context,
some look at meta data, but most combine more than one of those elements in some magic ratio that
is completely proprietary.
What that means for you is that if you design your web site for search engines, you’ll always be play-
ing a vicious game of cat and mouse. But if you design your web site for people, and make the site
as useful as possible for the people who will visit the site, you’ll probably always remain in all of the
search engines’ good graces.
Keywords and Your Web Site
04 1 9:35 63