2009 Search Engine Ranking Factors
At the recent 2009 Advanced conference in Seattle Rand Fishkin, of SEOMOZ fame, gave a very informative presentation on the current factors that have been determined as most significant in the ranking of websites in the Google SERPs.
SEOMOZ have a great reputation in the search industry for the quality of their data and some neat tools available from their website. Back in 2005 they published their first series of ranking factors, collated from information gathered from people in the SEO industry. This popular publication was updated in 2007 and looks set for another update with the latest information.
Although many of the fundamentals, highlighted in 2005 and 2007, remain the same there are some notable observations based upon correlation. SEOMOZ have not only sought feedback from notable people engaged in hands-on SEO, they have sought correlation from the search data provided by Google.
For example, although many active SEO’s say that the H1 heading tag is very important, data correlation suggests that it is far less important than is commonly believed.There are plenty of websites ranking very well for target keyterms and they are not making much use of the H1.
Contrastingly, the page title tag remains all-important. And keyword placement within the title tag is also very important with words closer to the start of the title having more significance.
Another point that is made by both feedback from the experts and data correlation is that onpage keyword density is not very important. It is noted that offsite, inbound linking factors carry far more weight than onsite factors such as keyword density.
Some estimate the relative split in importance to be around 70% applies to inbound link related factors and around 30% relates to onsite factors, including domain age. This clearly demonstrates the importance of inbound links.
One of the aspects of backlinks that remains important is the anchor text used in those links. The number of instances in which a specific anchor is used and the context in which those anchors are used is important.
Another point that is important to anyone engaged in link-building is that the diversity and number of domains providing backlinks is far more important than the number of backlinks. A site with 10,000 links from only one domain will be very weak in comparison with a site that has 10,000 links from multiple domains.
One interesting observation from correlation testing is that many of the sites that are ranking highly do not have valid markup. Many reputable websites don’t appear to be bothered by the need for validation. The basic advice was that as long as the cached version of a page matches what is to be presented to the site visitors then there shouldn’t be any problems from an SEO perspective.
What does this all mean for hands-on site optimisation? It clearly suggests that ensuring pages pass XHTML validation testing is not really very important, but checking the markup is still good practice.
Significantly, it highlights the importance of backlinks from multiple domains using relevant anchor text. So, to get a page to rank well for the term ‘Manchester Airport Parking‘ we would need to attract links from a variety of sources, each using this anchor text, or variations thereof.
Also, page titles are as important as ever so getting a page to rank for the phrase ‘Airport Parking‘ means using that phrase at the very start of the page title. Data also suggests that stuffing the page full of this particular phrase is not very effective.
Data also indicates that the H1 tag is far less important than had previously been thought. Whereas previously we might have insisted that a pages primary keyterm is used within the page H1 heading as part of the onsite optimisation, for example the term ‘Travel Insurance‘, we may be less insistent upon this requirement.
Related posts:
Guest Posts | seowhitehats | August 8, 2009
I think Link-assistant’s Website auditor and its pro tools done a great job in performing SEO.