Search Engine Optimization Tips

There is probably much that I have not included here, or that needs more clarification. If this overwhelms you completely, then it might be time to hire someone to do the work for you, especially if you don't have the time to do a good job. It might be fun to try to build your own deck, but you don't want it collapsing because you didn't know what you were doing.

What Search Engine Robots Look For

Search Engines bots look for text, plain and simple. Unless they are specialized, they are not going to be looking for animations, graphics, songs, programs or anything else. Why is this? This is because they cannot see, hear or interact. All they can do is read, and even with that ability, they really do not understand what has been presented to them. It is up to the search engine itself to search through its database and present the most relevant content.

Where do Search Engine Bots Look?

Bots look in certain places in a document first. Bots like things presented in a way that they understand. They especially like documents that have been marked up correctly. Imagine you are skimming a page with 2,000 words. Would it help you if the page was divided into sections, with heading clearly marked? Of course it would. It would also help if there were a title to the page, so you would know what the page is about.

Search bots like that type of formating, too. Therefore, markup your documents using Title element, use Heading elements, use Lists for lists, tables for tabular data. Do not use Heading elements to present text in a larger size. Do not use tables for layout purposes. The page should make sense structurally.

Let's look at the following example:

<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN"
    "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd">
<html xmlns="http://www.w3.org/1999/xhtml" lang="en">
<head>
<title>Botanical Glossary</title> Document Title
<meta http-equiv="Content-Type"
content="text/html; charset=iso-8859-1" />
<meta name="description"
content="A Botanical Glossary of Herbs and their Uses" />
</head>
<body>
<h1>Botanical Glossary</h1> Heading 1
<h2>List of Herbs and Their Uses</h2> Heading 2
<dl> Definition List
<dt><dfn>Medicago sativa</dfn></dt> Definition Term
<dd> Definition
<h3>Alfalfa</h3> Heading 3
<h4>Products</h4> Heading 4
<ul> Unordered List
<li>Product 1</li>
<li>Product 2</li>
</ul>
<h4>About</h4> Heading 4
Contains an antifungus agent.
<h4>Uses</h4> Heading 4
alcoholism, anemia, antifungal, arthritis
</dd>
</dl>
</body>
</html>

Will look something like this:

Botanical Glossary

Botanical Glossary

List of Herbs and Their Uses

Medicago sativa

Alfalfa

Products

  • Product 1
  • Product 2

About

Contains an antifungus agent.

Uses

alcoholism, anemia, antifungal, arthritis

Notice how simple this looks, one can easily see that this is Botanical Glossary, a List of Herbs and Their Uses, grouped by the common name, then by Products, About and Uses. It's all very clear, isn't it? Since this is a glossary that is being marked up, we have marked it up as a Definition List, and have used an Unordered List to show the list of products. We can use style sheets to change the presentation and make it look different, as shown in the Botanical Glossary at Intra Products using the float:left property to layout the definition horizontally.

Why is validating documents important?

Bots are only machines, and not very bright either. They are not fault tolerant browsers. If the markup is not valid, or just a bunch of tag soup, they may think it is garbage and go elsewhere. That is why it is important to validate all your documents, fix any errors, and then let the bot have its way with them.

If you were to copy and paste the above code as a new document, minus the highlighted text, and ran it through the validator at the W3.org, you would see that it would validate. If you checked off Outline as well, you would see a perfectly good outline.

It's like using a robotic lawn mower to mow the lawn 1. Here's what the robot will do:

  1. If there are no large rocks, then the robot finishes quickly and puts itself away.
  2. If you know that your lawn has a lot of rocks on it, then you have to pick up the rocks before you let the robot mow, otherwise,you'll mess up the mower.
  3. If you just let the robot go, and the lawn has a lot of rocks, then the robot will try to go around the rocks, but it won't do a good job. If there are too many rocks, the robot might just give up and put itself away.

Search engine bots like clean lawns without rocks. If you validate before you publish, the bot will not have problems.

Ok, I'm convinced — how do I do this?

Actually, learning to write valid markup is not that difficult. It may take a little time, but as my mother used to say, "If you're going to do a job, you might as well do it right the first time." So here is what you need to do:

  1. The first thing you want to do is sit down with a piece of paper (yes, that stuff they make out of trees), and make a general outline of what your site it going to be. Think of each heading as a page, then break the outline down farther for each page. At that point you will also have a reasonable idea of how long it should take you, and how many, if any images you want on each page.
  2. Now that you have framework, it is time to do the actual work. First, you need to get the right tools for the job.
    1. I suggest you get a good text editor to do this. Do not use a WYSIWYG editor, especially not Microsoft Front Page™. HTML-Kit is a great text editor with color coding, plug ins and is free.
    2. Because you are not going to be using tables for presentation (keep saying that - I am not going to use tables, I am not going to use tables), you should also have a good CSS editor. You can download a trial version of TopStyle, which integrates well with HTML-Kit and other editors.
    3. Have an HTML reference available. I think the best one thing to do is to go to the source, so you can download the HTML 4.01 Specification, or keep a window open to the HTML 4.01 Index of Elements. Another great source is Index Dot HTML.
  3. Now that you have the tools, it's time to start writing the content.
    1. Use a DTD. That should be the very first thing on your page. It tells the browser what type of document it is, and how to parse it. You also must have it to have the document validate.
    2. You should also include the Meta http-equiv, which tells the browser what type of document is being served up, and what character set is being used. You should do this, especially since some servers do not automatically do this. The validator also needs this information.
    3. Make good use of the Title element. Don't just call it Home. The Title element should be rich in keywords that are relevant to the document.
    4. Don't worry about the Meta Keyword too much. Because of search engine spamming, few robots pay any attention to it. However, you should use the Meta Description because some bots will use this in addition to the Title element.
    5. Write the body of your document, making sure that relevant keywords are in the content. Write grammatically, and check your spelling.
    6. Save your work. When you name documents, try to name them something that you can remember and that makes sense. For example, a botanical glossary could be called glossary.htm. Personally, I like to user lower case when naming my documents. It's a little easier on the Shift button, and you do not have to worry about Unix servers that are case sensitive.
  4. Time to do some testing.
    1. Validate your document. My suggestion is to use the validator at W3.org, then fix whatever errors it comes up with. If there are no errors, then pat yourself on the back, you did a good job. But, you're not done.
    2. View your page in various browsers to see how it looks. Don't think that Internet Explorer is the only browser - there are hundreds. Here a list of major players:
      • Internet Explorer
      • Netscape version 4.77
      • Netscape version 6
      • Netscape version 7
      • Opera
      • Lynx
      • Safari (Mac)
      • I-Cab (Mac)
      • Konqueror
    3. Optionally, download and install a speech browser. The IBM Home Page Reader or Simply Web are good choices. Listen to your page, see if you can navigate it with the monitor turned off. You can get a good idea of what a robot experiences.
  5. After your documents have passed the testing phase, it is time to publish. Put your documents and their associated images and other files on the server. If you have put different documents in different directories, you should use the same directory structure on the server.

To Submit or Not to Submit

Very often there is no need to submit your site to search engines. If there is a link to your document on a page somewhere, the bot will follow the link to your page.

That being said, if your site is brand new, and no one knows about you, then it is a good idea to submit. If your site is complex, you should have a site map, and submit that.

Where Do I Submit?

Robots will generally find you. Directories like Yahoo and DMOZ do not use robots and must be hand submitted. The best thing to do is the find the category where your site should be, and submit it there. Be honest. Don't try to submit to categories that have nothing to do with your site. Actual real live humans will visit and see if you should be included or not. Some of the directories charge for fast inclusion, DMOZ does not, but takes much longer for someone to come round. Be patient.

Search Engine Submission Pages
Search EnginePage Url
Googlehttp://www.google.com/addurl.html
AlltheWebhttp://www.alltheweb.com/add_url.php
AltaVistahttp://addurl.altavista.com/addurl/new

For DMOZ, or the Open Directory Project, go to DMOZ, find a category that fits your site, business, product or service, and submit from there.

Friends

Probably the hardest thing to do is to get reciprocal links. It is probably the most important thing to do as well. If you provide good content, a lot of sites will gladly link to you, even if you don't ask them to, or even if you don't link back to them. Look for sites that are similar to yours, whose visitors could use your product or service, and ask nicely for a link.

Here's the Don'ts

Outsourcing

As I said in the beginning, if you do not have the time to do a proper job, hire someone who can. Make sure who ever you hire is not trying to sell you a bill of goods. Don't hire someone who promises you a lot of shiny, flashy things. Hire someone who knows what validation is, because that person probably knows how to write valid markup. Ask what kind of editor the person uses. Ask for sample of their work, and submit their pages to a validator. If they don't pass inspection, don't hire them.

1 I am not kidding, as of July 2003, there really are robotic lawn mowers. Search Google for robotic lawn mower, and you'll find over 3,500 results.