Search Engine Optimisation

Search Engine Optimisation Primer

There are three main things you have to do to get your website a good ranking on the search engines (i.e. Google, which is the main engine everyone worries about.)

  • Technical issues: how the HTML is written, how the URLs are built – will the spiders like the page?
  • Textual content: is there enough, how often are targeted terms mentioned?
  • Linking: links in = good.

Technical SEO

Google and the other search engines use programs called ‘spiders’ to read web pages and process them in to an ‘index’ which lets people search on them. Spiders only care about the textual content on your page. If you make your blocks of text easy for the spiders to find and reduce the amount of extraneous tags and code on your page then they will like your pages more. So, using CSS to control the layout of your page will generally use less HTML on the page than using tables for layout does. Less code on the page means less clutter, which means the spiders disregard less of your page to get at the content. Also, there’s no chance of parts of your content being split up by code when they should be together, i.e. two paragraphs which are read next to each other, but have lots of table tags between them within the code.

Pages that are light on code are preferred by the spiders, so CSS can be a real benefit. You should always put CSS and Javascript in external files and call them in to the page, don’t put the code on the page itself as again it waters down how important the text seems.

Use semantically correct mark-up: for headers use H tags, for important words wrap them within bold or emphasis (em) tags, then make them look how you want using CSS. The indexing process knows that words within H tags are more important than other words on your page, so if your targeted phrase is in a heading and you mark that up with an h1 tag, the search engine will know those words are very important compared to everything else.

If you have a dynamic URL, especially ones with a session ID (the stuff that looks like ID=u324098a098q45u7 or similar) then the spiders see this as much less appealing. Generally if you have more than two variables in your URL the search engines will give it a very low ranking, if they bother spidering the page at all. Pages with session IDs tend to get ignored completely because spiders can get caught reading the same website time and again as new content when the session ID (and to it, the URL) changes. Therefore to stop this happening the spiders ignore anything they think might be a session ID, even if they’re mistaken and it’s just that your site uses something that looks similar.

To get around this problem, lots of people use ‘mod_rewrite’ to change URLs that look like index.php?blah=123&wibble=quack in to URLs like index.php/blah/123/wibble/quack. The spiders prefer this sort of URL as it looks normal to them and they don’t notice you’re actually running a dynamic site.

A utility which will read your web page and tell you what’s wrong with it is: http://www.spidertest.com (disclosure: written by the person who created this page.)

Textual Content

You should have at least 200 words on the page you want to rank well. Don’t try and target too many search terms on one page – choose two or three related terms and create more pages for the other terms you want to target. Add more pages tailored to specific topics rather than trying to put all topics on to one page.

If you can’t think of enough about a term to write 200 words on it, perhaps you shouldn’t be targeting that term. Alternatively, have a while to think about it and try and come up with different angles of talking about the topic, you might be able to think up more useful text to write.

Don’t forget people have you read your pages as well as search engine spiders, so make it legible to humans. It can be well worth hiring a copywriter experienced in SEO text to sort out the text on your site.

Links

Google likes pages that have links to them. To get your web pages high up for your targeted search terms requires them to have links from other websites to those pages. As this has been part of Google’s success at bringing up relevant results, the other search engines have copied this method of ranking and use it to greater and lesser extent in their own assessment of your pages.

Preferably, have your targeted search term as the text of the link if possible.

Google also ranks how worthy each link is. If a link to your site comes from a site that itself has lots of links to it then that link gives your page a better weighting than a link from a site that no-one has linked to. This starts getting in to what Google calls ‘Page Rank’ – a scoring out of 10 for each web page it knows about. Getting a link from a page with a Page Rank (PR) of 6 is more valuable than one from a PR of 3. To find out the PR of your own pages and of people who are linking to you, the easiest way is to download the Google Toolbar for PC Internet Explorer from http://toolbar.google.com and turn on the little indicator of Page Rank. This gives you a bar graph of the rank of every page you visit, letting you know how useful a link from a site is, and how well your own site is doing.

Linking is a very important part of SEO and if you want your pages to rank well, it’s imperative that you spend time getting links built up both to the home page of your site and the most important pages within it.

Things people think are important, but are not

The meta keywords tag is now basically ignored by search engines because it has been so abused in the past, i.e. lots of websites with nothing about Britney Spears still had her name in their keywords tags to try to encourage visitors.

The meta description tag should be set individually for each page on your website. Do not just put the same description on every page of your site. It’s not there to describe your website, it’s there to describe the individual page it’s on. The description is read by various search engines, and is sometimes used as part of the description of your page on the search results, so it should be useful for a human to read as well as spiders.

There are rumours that Google does read meta keywords and description, not to add them to the words your page targets, but to look and see if they match the content and themes on your page. If they do not it gives a slight penalty to that page. This is partly why it’s important to fill them out for each page of your website, not just having a generic description that you use everywhere. Filling out individual descriptions is not a glamorous part of the job, but you could be damaging your site if you do not do it.

How to get banned from Google

If you want to get your website dropped from Google, which is kind of like not being on the web at all these days, you can try these naughty techniques:

  • Hiding content on your pages, e.g. white-on-white text, using CSS to position a DIV off the screen
  • ‘Cloaking’ – showing different content to the spiders than to human visitors
  • Link Farming – having pages and pages of links to places you are trying to promote

You may come up with a technique that Google doesn’t ban automatically. Unfortunately, a competitor, or their SEO company, will use Google’s spam report form to dob you in to Google and get you banned – a very easy way of getting their company’s results a bit higher.

If you use an SEO company and you think they are using any of the above methods, you should get someone else to check your site out. If you ask on the Freelancers.net mailing list someone will be able to have a quick look and see if anything bad is going on.

further reading