Accessible content
3. Be careful with images. Use CSS Image Replacement
Everyone by now is (hopefully) always setting the "alt" tag on any images they include on a page. However, search engines still treat these differently to straightforward HTML on the page. Following on from point #2 above,
<a href="graphics-card.aspx">Our Graphics Cards</a>
is generally weighted much more than
<a href="graphics-card.aspx"><img alt="Our Graphics Cards" src="graphics.gif" /></a>
If there are images on your site that aren't content in themselves (such as navigational links and headers), then you should be using a CSS Image replacement technique. This allows to you do something like this:
<a id="graphics_cards" href="graphics-card.aspx">Our Graphics Cards</a>
/* css */
#graphics_cards{
display: block;
text-indent: -5000px;
background: url(graphics.gif);
height: 25px;
width: 25px;
}
This technique can (and should!) be applied to headers too, if you're using images there.
4. If you must use flash, put some text in too
Using something like SWFObject allows you to embed a flash movie, but also include a HTML equivalent that will be indexed by search engines, and viewable by users with flash or javascript disabled.
6. Ensure content is accessible via hyperlink (watch out for AJAX!)
Sounds simple, but if you're using any sort of form submission (such as a drop down box) or ajax-style scripts to navigate to content on your site, then your content is not going to be indexed unless you take a few further steps.
a) Ensure the content can be accessed directly from a distinct URL too (ideally in conjunction with url rewriting - see point #7). If you've got a drop down box listing frequently asked questions, when an item is selected, redirect the user to a unique url for that FAQ entry, rather than just post-ing back and adjusting the content of the page.
b) Ensure you've actually got links pointing to these distinct URLs (whether within elsewhere on the site, or externally). You could also use a Google SiteMap so that at the very least the search engine knows those pages exist.