• 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

Do you understand what your Web pages say?

Posted by Michael Martinez on May 13, 2007 in SEO Theory

There was a time when Web documents were constructed primarily with the user in mind: the person reading the content came first. And then we decided we had to get fancy, start adding stuff to the presentation to make Web documents look more like printed documents.

We added pictures, tables, embedded objects, frames, and other whizbang things. But to add this stuff we had to resort to special instructions intended more for the browsers than for the people.

To simplify the task of talking to browsers we added a styling language (CSS1). It wasn’t capable of much but we added layout features in the next generation (CSS2).

We even began telling browsers to stop using their own gizmos and use our own (Java, Javascript). We found ourselves having to tell browsers to stop misbehaving (MS Smarttags). We even had to prevent browsers from accessing our pages as if they belonged to other sites (frame-busting Javascript).

And we found that we could also talk to our servers from within documents, asking them to include content from other files that browsers cannot reach; asking them to serve new files after people take specific actions; etc.

Somewhere along the way we also began talking to search engines, asking them to follow, index, archive, not follow, not index, not archive stuff. And now we can tell them not to use directory descriptions in their search results pages, not to follow links on a link-by-link basis, and even to ignore certain sections of our pages that our visitors still see.

The end-users are almost completely forgotten in the cacaphony of commands being issued to servers, browsers, and search engines. Web document standards have failed miserably to separate the engine from the interface. Quite the opposite effect has happened: the Web document, which was originally just the interface, is now the engine driving all the parts.

Search engine optimization therefore has to take everything into consideration: users, servers, browsers, and search engines. Is it any wonder that SEOs stumble around the Webscape blindly poking holes into their own strategies?

We now have to remind ourselves to write copy for users, not search engines.

We use our servers to create virtual Web sites, grabbing content from hidden files and SQL databases, formatting pages on the fly.

We have to tell browsers and search engines alike to substitute URLs, to stay out of directories, or to go away completely.

And we have to tell the search engines to crawl our pages.

It was never enough for a Webmaster to just let the technology do what it was supposed to do. We had to take control wherever possible, and each time we took control over the experience the other side fought back.

Users left our cheesy sites in droves, forcing us to improve our design and content.

Servers buckled under the demand, forcing us to implement new technologies and buy more powerful hardware.

Browsers choked when rendering our convoluted pages, forcing us to advise users to surf with our favorite browsers, or to create two or more sets of content pages for those other browsers.

And search engines looked for creative ways to keep our cheesy pages out of their indexes, or to suppress our lazy content because there was something else out there which was less convoluted, more useful.

The lesson history wants to teach us is that challenge increases with complexity. That is, the more we try to do with our pages, the more difficult it becomes to do anything well, much less achieve the results we want.

Most people will read a page that has nothing more than a few navigational links, a header, and several paragraphs of text. Nonetheless, we insist on inundating them with autoloading videos, interactive gizmos that take control of the browser away from them, and hard-to-read fonts.

Most Web content can be presented in a straight list format, top-down. But now we struggle to design cascading style sheets that can take the linear content and make it appear to be spread across a page as if it were laid out on a paste-up board — and we’ll gyrate through any amount of bloated CSS code in order to avoid using HTML tables ( which were originally used for page layout because we didn’t have any other way to do it).

It takes less effort to create static HTML pages than it does to write a system that generates them dynamically. But we invest more and more time and resources each year into developing ever larger, more complex dynamic content generation applications.

And search engines don’t need anything more than the content you put on a page to understand what it is relevant to, but we devote an increasingly lop-sided amount of time and effort to building “relevant” backlinks as well as to embedding backlinks that won’t pass useful anchor text.

If you conduct a poll asking every SEO you meet what two factors are the most important for controlling search engine rankings, how many do you think will choose page titles and links? How many do you think will choose page titles and something else?

A well-rounded SEO has to understand how .htaccess files work, what server side includes are, the difference between sub-domains and top-level domains, robots.txt, which meta tags are used for which purposes, and how cascading style sheets both simplify and complicate the page layout and design process.

That same SEO needs to be aware of what on-page emphasis can do for asserting relevance, and also has to know what to look for to discover if a linking partner is cheating in an exchange.

Our competent SEO needs to have enough common sense to know that if crawlers cannot fetch pages during server down time those pages may not be indexed. Our SEO also has to understand that duplicate content may make it difficult for a search engine to decide which page to show, and that inconsistent URL references in links may cause search engines to see duplicate content that doesn’t really exist.

A well-trained SEO has to think about how a page will look in Ask, Google, Live, Yahoo!, Internet Explorer, Firefox, Opera, and Lynx. The same SEO knows copy needs to be compelling as well as informative. It has to be easy to read, easy to understand, and easy to follow.

There are a thousand little things you need to know and understand. But the most important thing you need to know, the most important factor for controlling search engine results, is that your Web pages must be multi-lingual, fluent in all their languages, and as simple as possible.

You should try designing a Web site that uses nothing but a handful of HTML elements, maybe only H1, OL/LI, and CENTER (allowing for links between pages). You might be surprised by what you’ll learn from returning to the basics. Your pages will speak to you in a very clear, elegant way.

When you learn to listen to the rhythm of the Web documents you create you’ll be better able to diagnose their problems. At all levels of design and optimization you should seek to achieve symmetrical structure. Anything that breaks the symmetry disrupts the rhythm.

And when the rhythm is broken, your page will speak to you. You just need to listen, hear, and understand what it says. That only happens at the highest level of search engine optimization. That should be your goal.