Nathan Buggia is up on stage again. He starts by disecting from an SEO standpoint a website called Arbor Snowboards. The site is very flashy and interactive, but when you look at the source code, the meta title alone does not say anything about the site — meta title is “Arbor: Home”.
Unsurprisingly, while the site ranks number 1 for the term “Arbor”, it ranks 49th for the keyword “snowboards”, which basically what the site is about. Nathan suggests ways to optimize the code. Some AJAX and HIJAX best practices:
- Learn about Hijax at http://janeandrobot.com/admin/Pages/seo-developer-resources.html.
- Use <noscript> and <object> tags.
- Don’t use JavaScript for links.
CSS can improve performance by separating formatting from content. However, it has been abused by Spammers to try and hide links and keywords. (I.e. setting fonts to 0, or setting color to #FFFFFF).
Use <h1> instead of, say, <span class=”heading1″>. Use HTML tags semantically.
Sharad Verma, Sr Product Manager, of Yahoo@ Search, gives a picture of Web 2.0 to start with. He shares that there are 3 pillars of Web 2.0:
- Experience (last.fm, youtube.com)
- Participation (tagging, reviews, comments, wiki, social bookmarking). Users are more engaged in the web now.
- Community. (Are you on Facebook?) People are uploading and sharing stuff on the web.
CSS is good, external CSS is even better. Yahoo! understands content layout. Through CSS, you can rank for your keywords better and place important stuff higher in your codes. It promotes efficient crawling. It makes pages light.
Don’t disallow CSS in robots.txt. Yahoo! would like to understand your layout also. Don’t hide text using CSS.
Is your site a website or a web application? Some sites are actually web applications. They are functionality driven, don’t need to be deep indexed. Web Sites, on the other hand, are content driven, hence need to be deep indexed. Use AJAX selectively.
Spiders can’t parse AJAX content. AJAX is delivered via JavaScript that search engines have difficulty crawling and indexing. Spiders can’t extract java script links. AJAX content can’t be linked to, can’t be bookmarked.
If you have to have AJAX, make sure that your initial page should have content and navigation in html. Use href links. Turn off JavaScript in your browser and ensure that the content is navigable. Be careful with mirror sites. If your AJAX site has an HTML version, the HTML is the one that will be indexed.
Yahoo! indexes flash. The problem with flash, however, is that they are rendered via JavaScript. Because there is no deep linking, there is less link juice. More over, one URL for the entire flash movie equates to bad user experience.
He ends by promoting the new site explorer, which is now out of beta. It has new site features, allowing you to submit sitemaps, delete URL paths, and rewrite URLs.
Daniel Pang, Software Engineer of Google, starts by discussing also the web 2.0 environment. Much of Web 2.0 is already search engine friendly. Web 2.0 SEO is more designing the site for user accesibility and experience, rather than for search engines. Does your site work in Lynx? If it does, people will be able to use it everywhere (online, mobile, etc.)
Web 2.0 Flash / AJAX. Web 2.0 sites are now embellished with Flash and AJAX elements of richer media (ie. Flickr and Youtube). Daniel discusses a bit of HIJAX also (this I haven’t heard of until now).
Googlebot can read some content and links within flash, but not all. Best practice is HTML for content and navigation, complement with flash for applications and animations.
CSS allows one to separate content from presentation. It’s great for making changes to visual layout, and is generally not much to think about specially fo accessibility/search engines.
Side note: I didn’t know Danny S could be so soft spoken. :)
Tony Adams, SEO Manager of Yahoo!, is up next. He says that Web 2.0 is all about “sleek” designs. A dilemma for search engines, however, is that content is often controlled by JavaScript which is not search engine friendly.
Be strategic in the AJAX implementations. Separate content from presentation and interactions build using scripting language. Work with standards or platform teams to build standard interactions.
(This I have to try) Web Developer Toolbar Firefox addon. It allows you to disable Javascript, CSS, images, etc, allowing you to actually see what the search engines are seeing, on the site content standpoint.
Q&A. Danny Sullivan starts of Q&A with a question on H1 tag, if there’s value in setting the keyword to an H1 tag, even if you set it to look like normal text in your CSS.
Sharad: H1 tags don’t seem to be an important factor.
Nathan: He would question scripts that make H1 tags styled to look like everything else.
Daniel: H1 tags still indicate logical labels, especially when there’s no CSS and AJAX, so there’s still value to it.
Regarding sitemaps:
Daniel: Sitemaps is supplemental, but its better to have more internal links.
Nathan: In making sitemaps, don’t put all URLs, just put the best URLs there. The sitemaps are used in identifying which pages are most authoritative.
On URLs:
Daniel: Better to use nice, clean, and short URLs. If your URLs have a lot of slashes, rename them to shorter versions. Don’t just replace ?s and #s with slashes because to them, “/” does not translate to separation of paramenters.
(Okay, I suck at documenting/transcripting the Q&A, the panelists are talking a handful! I should just post this one.) :)