SMX Advanced: The New Periodic Table of SEO
By Tad Miller
Jun 7, 2011
More Articles by Tad
First Session of SMX Advanced 2011 in Seattle. The New Periodic Table of SEO, with Jeff MacGurn @yerrbo, Jamie Steven @jamies check that Rand Fishkin now pitching in, Matthew Brown @MatthewJBrown & Duane Forrester @DuaneForrester.
Danny Sullivan kicks things off talking about how what worked in 1997 doesn’t really work today for SEO tactics, it’s a history lesson of alogorithms and attempts at prevention of black hat tactics with them.
Jeff MacGurn from Covario starts:
The periodic table is not set in stone. It’s constantly changing. Covario tracked 800,000 landing pages as part of their study of ranking factors. Jeff brings up the “Correlation isn’t Causation” idea [smirk by Rand Fishkin]
- Page size, URL length and Flash Navigation not really correlated with ranking well , Dynamic Parameters in URLs not really correlated to ranking well at all
- Page Load time has surprisingly high correlation to ranking well
- Keyword emphasis doesn’t correlate well
- Keyword in Title Tag is Strong Correlation
- H1 & H2 Tags with keyword are strongly Correlated – not the H3
- Keyword in URL strongly correlated
- Keyword in Alt Image Text not strongly correlated
Links and Correlation:
- Internal Link Counts – little correlation to high ranking
- External link “counts” not much more correlated than internal link counts
- Keywords in Anchor Text is actually waning – but still slightly correlated
- Getting Links for Hubs – sites that link to other authorities was the STRONGEST correlation factors in the study.
Ranking factors change to fit the “market”. Algorithmic ranking factors change as everyone optimizes for those factors. Everyone that ranks for “smartphones” all have it in their title tag. Other factors have to be at work, because all the top ranking sites for that word are optimizing for it. Focus on differentiating factors to get ahead in addition to the basics.
Next is Rand Fishkin through the suposedly Poisoned? Jamie Stevens. Rand goes into SEOmoz’s correlation data:
See the latest and greatest here: http://www.seomoz.org/article/search-ranking-factors
SEOmoz looked at top 30 results and correlating factors for their study. It isn’t a reverse engineering of search algorithms. It’s just correlation.
In the SEOmoz survey, link-based factors have shrunk as an algorithmic ranking factor – according to the surveyors, and Social factors have grown.
Domain level link data is growing as a ranking factor and page level link data seems to be shrinking as a factor.
Exact match domain URL as a ranking factors has declined quite a bit in the last year (My personal opinion – it’s still a big factor from what I can see).
Social factors were huge in the survey data as being the newest correlations to rankings. Everyone said in survey that Twitter was the big factor – correlation data indicated that Facebook shares had the highest of the correlation to top 30 results of all the social factors. It’s a “Crap Ton” – a technical term of art. Rand has already written about this here.
Links are still the engine that drive search engine ranks. The other new factors are just being blended-in.
Brand-Social-Local factors are starting to be the factor that is shaping search results.
Brand = Trust. That’s entwined within Social. Google is guessing that intent for “Microwave Ovens” is transactional and giving brands the benefit of the doubt. Clickstream data gives search engines the “Brand Signals”. Natural link profiles always show brand keywords as the top anchor text links.
You need to build brand and social signals to your site. Facebook Open graph buttons are a great way to do this. Use open graph tags on objects and images. LinkedIn shares and tweets seem to correlate well to rankings for them. You need to figure out what “buttons” you need to prioritize based on your results and your audience.
Local results are even pushing Wikipedia down the SERP pages. The Google 7 pack and Bing 3 pack are starting to dominate SERP pages.
Try using Proxify to see local results and examine how localization is impacting your results. Thousands of Geo-specific pages will get you smacked by Google Panda.
Don’t wait for the new factors to become new factors, you need to “beat the market” to new factors like Google +1.
Duane Forrester from Bing – says don’t stray to far from the concept of “Content is King”. User experience is a factor in ranking on Bing. The crawler is a user as well as the searcher. Compelling content (especially with Schema.org tags) will be a factor. The technical stuff is fine the content is what matters.