• Register
• Log in

SEO Theory - SEO Theory and Analysis Blog

Algorithm analysis, Web community relationship analysis, SEO practices and techniques, industry news, etc.

• Home
• About SEO Theory
• SEO Glossary

SEO by the book

Posted by admin on January 23, 2007 in SEO Theory

Many companies have literally had to write their own SEO manuals for in-house reference. There are, of course, plenty of search engine optimization books you can buy. Some of them sell very well. But the undisciplined nature of the search engine optimization industry virtually guarantees that most (if not all) SEO books include a great deal of misinformation, myth, and nonsense.

It doesn’t take much to derail a potentially useful resource. All you have to do is say or imply that links are the key to search engine optimization success. That bit of nonsense is repeated every day in forums, FAQs, and tutorials, so most people conclude that it must be true. No one has ever proved that links are key to search engine optimization success.

The only way you can prove such an assertion is to show that any strategy that doesn’t depend on linkage will fail. That’s impossible because there are many competitive queries where linkage doesn’t determine the results. Most people make the mistake of assuming that just because you can rank by links that must mean ranking by links is the only way it works.

And what is sad is that ranking through linkage is the hard way to do it. After all, most Web pages are not optimized for relevance. So if you drop an optimized page down into a field of unoptimized pages (and no one is trying to rank by links) then links aren’t a necessary ingredient for search engine optimization success.

Hence, the idea that you must rank by links is false. It’s a myth concocted by incompetent people who never learned how to get rankings any other way.

Now, some people are quick to point out there are queries where you pretty much have to rank on links. After all, everyone else has invested in the linkage to ensure their success, so in order to compete you have to fight them with links. Well, yes and no.

For search engines that allow links to pass value, links help with rankings in two ways: they pass anchor text and they pass a quantified value we can call PageRank or Link Popularity (which are two distinct concepts). Neither PageRank nor Link Popuarity can affect relevance in any way. Anchor text can.

So links can help by making a page more relevant and by making a page more important. And a few years ago it was so easy to use links to rank that most people stopped caring about the other (easier) things you could do to achieve high rankings for many competitive queries. But the search engines began filtering links, and gradually fewer and fewer links have proven to be influential in affecting search engine results.

So, yes, if three years ago you could have scooped up 1,000 links in a short time you were good. But do those 1,000 easy links count today? Most likely not, although it’s true that some worthy content which was propelled to the top of search results with artificial links has by today actually earned natural linkage.

Before link building became a spammers’ playground, before it was easy to go out and get a hundred links, people used to struggle with getting links. Back then we used to advise people to select links for the direct traffic and visibility they could create because often the easiest links they could get weren’t likely to be noticed by the search engines. Those were the days when Inktomi ruled the Web and it ignored most links.

Now things have come full circle. Google rules the Web, it ignores most links, and I’m starting to see more forum moderators and admins advise people to get links for the traffic and visibility they will create. People don’t realize it but pursuing links for these reasons rather than for search engine success will actually improve the overall quality of inbound linkage.

In Web site marketing, deprivation is the mother of alternative sources of traffic. Once you lose that crutch you call Google, you can either fold up your Web sites and go home or you can wake up to the fact that there are other ways to get traffic. Sure, you can buy PPC ads from Google but eventually you’ll get tired of paying for traffic unless your return on investment is huge.

The art of on-page optimization may or may not make a comeback. If it does, it will be in spite of all the bad and incomplete advice you’ll find on the Web. There are honestly very few people who know how to do it well (from a technical or mechanical perspective) because the vast majority of people arbitrarily dismiss all the information Google has provided through the years.

There is little if any consensus among SEO thought leaders about what constitutes useful on-page optimization. Nearly everyone agrees it’s a good idea to put your keywords into the page title but I’ve actually seen a couple of SEO forum operators insist it probably doesn’t help much. Not that they’ve ever tested page titles that I can tell, but there are reasons to agree with them (Google claims over 100 factors are used) and reasons to disagree with them (many people do actually top out rankings solely because of their page titles).

After the page title, almost everyone recommends using a meta description tag but you won’t find much agreement on whether it helps with relevance (it doesn’t).

Then you have the keywords meta tag, which actually still helps with relevance but most people say it’s now useless. But include it anyway, perhaps out a gut feeling or some SEO superstition that makes them feel better.

Why do I say that meta descriptions don’t work and meta keywords still do? I’m not going to tell you. Most of you could find out quickly enough, but have you bothered to try? What do I keep saying what constitutes search engine optimization? Experiment, evaluate, and adjust. I actually do that. I don’t just sit on forums and repeat the same nonsense endlessly without checking it every now and then.

But most SEO guides pretty much stop here. Oh, you find a few people who talk about using keywords repeatedly throughout the page, and some who talk about Hx header markup elements, and some people even talk about using bold or strong. But most on-page optimization factors are almost never mentioned. It’s not like these factors haven’t been published by credible, authoritative sources.

And the trusted experts cannot find any consensus about the value of the on-page factors they may actually recommend. They’ll insist most on-page factors no longer matter much. They have no sound reason for saying this, except that their friends and fellow experts are all telling everyone that you have to build links in order to control rankings.

SEO testing is another area where people make critical mistakes. The most common mistake I see is the resistance to invasive SEO. That is, testers have convinced themselves that tests need to work with unused keywords. Now, I have actually used previously unused keywords for some tests, but that doesn’t mean all tests need to be performed this way.

In fact, some test can only be performed in competitive queries. After all, how can you determine if something actualy works if you avoid trying to use it where it needs to work? That makes no sense. In order to force a search engine to sort query results according to its normal criteria, you need to feed it more than 1,000 pages in order to get it to a point where it has to throw everything it can into determining relevance.

Otherwise, the search engine can take your 5 test pages and sort them on the basis of which page was crawled first, which page loads fastest, which page has the shortest URL, which page has the alphabetically shortest title, etc. That is, the search engine can sort items much faster when it has fewer items to sort, and it doesn’t have to be as picky. After all, if you only have 5 results for a query, the order in which they are displayed pretty much doesn’t matter. Most people are as likely to click on all 5 results as not to click on any of them.

But the best reason not to use previously unindexed keywords for most testing is that the time you take to create your test pages can be used to create real content pages that have some long-term meaning and purpose. So what if your first attempt at optimization doesn’t work? That’s what search engine optimization is all about.

You experiment, you evaluate, you adjust.

That’s really all any SEO book really needs to stress. Everything else is just Web design, site promotion, branding, and market research.

Those topics are a little more complex than mere search engine optimization.