Google Onsite Optimization – Likes and dislikes
So here we are: the first session of the A4uexpo 2011 in Munich. Sepita Ansari of Catbird Seat. Moderator Alexander Holl introduces Sepita as “a very famous German SEO”. His topic “Google onsite optimization” could be very boring, but as I know Sepita as a good entertainer it will be fun!
SEO is a black box. You don’t know what will going to happen when you make a change. Sepita shows the periodic table of SEO ranking factors of Search Engine Land. He’s going to talk about the left part of the table: the onpage factors.
The Panda update is about thin content. In the next few months the Panda update will probably be rolled out to non-english speaking countries.
But what is thin content for a machine? Sepita differentiates between on page factors, content factors and bounce rates.
First of all you have to understand the user intent. You have to answer the question the user asks by the search query input.
1. Keyword research
2. Offering the right content / urtl
3. Tying the user to your site
ad 1. Content research & keywords
- Basis of SEO
- Continuous process
Proper keyword targeting:
- Head terms – mid-tail -long-tail
- High value words: (Sepita suggests to read this post of Bill Slawski)
Pitfalls are cannibalization and formatting. Use a proper formatting. Some SEO’s just put in plain text, without a good formatting. Google will understand that you are just putting in the page for SEO purposes!
ad 2. High Quality content:
- Signficiant unique content above the fold
- Write good articles: Introduction-middle-conclusion
- Linkable content
- Homepage vs. landing page content:differentiate between a general content on the homepage and specific lp content
Pitfalls are duplicate content caused by the CMS used and filter / pagination duplicates.
ad 3: Tying the user
- bouncing issues -> landing page optimization
- suitable pictures
- suitable videos
- user generated content
- ratings in combination with microformats (SERP rich snippets)
- call to action
- Query deserves freshness, Google will crawl your site a lot more when you have fresh content.
How to avoid content troubles?
Avoid short content, if it’s not ranking just put a noindex or even a 404.
Duplicated content (intern and extern), redirect it
Low unique ratio on page (template content, with just small differences)
High Ad ratio
- automated content or templated content
- copied or poorly spinned content
- spelling mistakes, run a spell checker, because Google will!
- optimizing on semantic close keywords
- search in search
Now Sepita continues with some tips against a Panda problem:
- use trust seals
- high value words from an expert editor
- spell check program
- use analytics to find out zero-traffic pages
- social content seeding
- engage the user (via pictures, lotteries, games etc.), so your user will not bounce.
In the second part of the presentation Sepita will talk about some HTML rules to follow. I think we all know them, but many people are still doing it wrong.
The HTML titles are the most important tag of course. Sepita strongly suggests to make them at least individual on the most important pages. They should:
- match the intent of the page
- content match
- be a descriptive marketing sentence
- be 65-68 char.
Pitfalls are keyword repetition or making the titles too long
The second important tag are the description. Also the description should be unique for the most important pages. Besides that descriptions should:
- be a descriptive marketing sentence
- animate to click
- improve CTR
- be 120 – 160 characters (incl. some keywords)
Descriptions are not a direct ranking factor.
The biggest pitfall: repetition of the title tag.
HTML – headers:
- Unique headlines including keywords
- Structured hierarchy:
Pittfall: no structure
HTML in general:
- HTML sequence
- HTML semantics
- Internal links: navigational systems vs. contextual hyperlink navigation
- Internal links: horizontal vs- vertical links
- Good internal anchor text
- Too many (footer) links
- Check for broken links
HTML violation: stuffing / hidden
HTML noticeable problems:
- Keyword overuse in tags -> use synonyms. Sepita tells the story of someone getting kicked because he automatically wrote the title tag in the alt texts of his pictures.
- hidden links/ text in HTML code
- Google: read more (hiding text behind a “read more” link, making it invisible for users, but visible for Google)
- Google: HTML validation
In this final part Sepita talks about a good site architecture and some mistakes, people make.
1. Crawling: Index control & pass reputation:
- do a site: search
- use robots.txt wisely
- follow / nofollow for extern sites
- Index / noindex
- canonical -> very powerful
- XML sitemap
- https issue
- too restrictive indexing
Google wants to make the internet quicker. Some tips:
- Media: optimize your images / flash / videos / audios
- File: use just 1 css file, reduce http requests
- Web hosting
- Optimize your code
3. Readable URLs:
- Keyword focused
- Marketing perspective
- Lifelong URLs: don’t change your URL’s!
- Small letters
- URL handling
- Not a ranking factor.
- Keyword overuse
- Deep directories
- Dynamic URLs, Session IDs, unnecessary parameters
4. Structured Navigation:
- Logical directories
- Flat hierarchies
- Mass links
- Useful 404
- Navigational structure (people drop in ranking by keyword stuffing in the navigation menu)
- Deep hierarchy (landing page – clicks to content)
- Facetted navigation