Ok, here is a poorly formatted liveblogging of this session:
Maile Ohye of Google sez:
Talking about ecommerce issues where we have a site with 158 products but because of filtering there are 380,000 URLs, so Google doesn’t know what to crawl.
- Maintain a consistent URL structure
- Directories and filenames are case sensitive. http://apple.com/itunes/ & http://apple.com/ITUNES/ are considered to be different URLs
- Keeping consistent reduces duplication, facilitates more accurate indexing and simplifies your robots.txt configuration
- 301s & rel=canonical are crawled much less frequently than 200s
- 404/410 URLs are crawled less frequently
- 500 errors are treated as a transient error. Pages not removed from index. We will retry in the near future.
- Use standard encodings & key=value (e.g. /product.php?item=nexus-one&category=mobile) v. non-standard
- Crawlers interpret standard keys & values.
- Use the URL parameter tool with Yahoo & Google Webmaster Tools. Tells bots which parameters are relevant and which they can ignore
- Indexing priorities: Googlebot looks for what users will find relevant: URLs with updated content, URLs with unique/important content (as determined by linking signals); Sitemap info and bandwidth considerations
- How to increase Googlebot visits: Strengthen indexing signals via uniqueness & freshness. How well the page is lined from your site and other pages on the Web.
- Use proper response coes
- Serve content reliably
- Prevent crawling of unnecessary pages
- Optimize performance: Shopzilla improved conversions by 7-12% just by increasing site speed
- Improve long-tail content: unique & fresh content, get links to these URLs
- Reduce duplicate content: Choose canonical URLs and be consistent. Include the canonical URL in internal links and sitemap. Use 301 & rel=canonical.
- Include microformats & RDFa: Enhances results with rich snippets – ability to include reviews, recipes, people & events.
- Use Video Sitemaps
Adam Audette of Audette Media
It’s all about user experience. Users come first and then the SEO.
4 Big Issues with SEO & IA Right Now
- Categorization, Search & Browse: Amazon provides key categories on the homepage but as you click into categories you get relevant sub categories and links to important product URLs in the category.
- Make Use of Link Relationships
- Know Your Internal Link Profile
- Content is more important than ever
Great user experience but bad for bots
- Rewrite facets to pretty URLs based on priority
- Place faceted experience in a folder for more control
- Append “overhead” attributes (e.g extra parameters) to the pretty URLs; rel=canonical back
Image Search Signals:
Content signals (color, facila recognition, etc.)
Attribute signals (ALT text)
Textual signals (captions)
Quality signals (pixels, etc.)
Images inside of flash/js are hard for the bots to access.
Provide dimensions of each image in the href whenever possible
Use JPEG for photos (strip meta when appropriate)
Use PNG for graphics
Use GIFs for small and animated images
If you register image with Creative Commons you can add more data to your image.
Provide as much info as possible along with your image:
Putting images in keyword relevant directories will help (e.g. images/lady-gaga/)
According to Maile, the major signal for site speed is client side rendering