Welcome to #SMX Advanced 2011! The science of SEO. It’s a code everyone and their mother has been trying to crack since the dawn of organic rankings. Just when we think we know the elements to a success by way of SEO, something changes, wanes in significance, or gets added to the mix. The traditional chemistry between optimized titles, tags, anchor text, on-page content et. all are no longer the only players in the petri dish. The social side of the user experience is gaining some serious traction, and factors such as merchant reputation, user reviews, content quality and social signals are becoming increasingly vital to healthy SEO.
Matthew Brown, Founder, AudienceWise, Duane Forrester, Senior Program Manager SEO, Microsoft, Jeff MacGurn, Director of SEO, Covario, and Jamie Steven, VP Marketing, SEOmoz took the stage on the morning of #SMX Advanced 2011, ready, willing, and eager to tour the audience through The New Periodic Table Of SEO. Moderated by Search Engine Land’s Editor-in-Chief, Danny Sullivan, and Jeff Ferguson, CEO, Fang Digital Marketing, together we donned our proverbial lab coats and goggles, and saddled up to get scientific about SEO. Read on for the full scoop.
First up was Danny Sullivan, who began with a thoughtful look at the Periodic Table of SEO as it’s evolved over the last decade.
The Periodic Table of SEO: A Brief History
- Back in ’95-’97 – the old periodic table consisted of titles, crawlability, description, headers, URLs, numbers; KW stuffing was the big no-no.
- As things evolved, Overture & Google came along, offered up the opportunity to do keyword research. Cloaking became the big no-no.
- 2003, 2004 – Florida update rolled out – quality content becomes important. A site oughta have authority, history. No-no = link spam.
- 2007 – content freshness became more important – locality became more important. Paid links = no-no.
- By the end of 2009, everyone’s SERPs were personalized, whether or not you were logged into Google. Personal search history = very important.
- Nowadays… it’s all about social signals. Reputation, shares, authority, engagement, etc. Big no-nos = blocking (i.e.: if individuals block your content, a sign of irrelevance, etc.)
Next, it was time to explore the current state of SEO ranking factors. Kicking off the presentations was Jeff MacGurn.
Biggest takeaway from the get-go: the Periodic Table of SEO is NOT set in stone (not even the new one). It is & will always be continuously expanding. Second biggest takeaway: Jeff is the king of incorporating Internet memes into decks.
Jeff was ready to take us through Covario’s SEO Factor Correlation Analysis – a research project that studied 800k landing pages, 56 factors, and 937 data points between Nov – Dec 2010.
Technical SEO Factors – Elements & Findings
- Page size –> Not highly correlated with ranking well
- URL character length –> Not necessarily correlated to ranking well
- Flash navigation –> Not correlated
- Session IDs –> Strong negative correlation when present
- Dynamic parameters in URLs –> Not correlated
- Proximity of page to root directory —> Low correlation
- Page load time –> Surprisingly strong correlation
Content Factors – Elements & Findings
- Keyword emphasis –> Not well correlated
- KW in title tag –> Strong correlation, stronger on Bing & Yahoo than Google
- KW in meta description –> Not correlated
- KW in H1 & H2 tag –> Strongly correlated overall
- KW in H3 tag –> Not strongly correlated
- People don’t really use H3 tags that much, maybe that’s why it doesn’t correlate as much as H1 & H2
- KW in image alt text –> Not correlated
- KW in URL –>Strong correlation overall
Link Ranking Factors – Elements & Findings
- Internal link count –> Very little correlation
- External link count –> Marginally better than internal
- Keyword in anchor text –> Waiting, but still slight correlation
- Link hubs –> Very strong, one of the strongest factors of all examined factors
- Mantra: “I want to have links from everyone my competitor has links from.”
Top takeaways: In the end, y’all can have the most amazingly optimized site ever, but UNIQUE content, a unique factor, is the most what will make all the difference, i.e.: Don’t look for the secret SEO factors of the day – the landscape changes without the fundamentals changing. You need to find other differentiating ways to rank well.
Next up, Rand Fishkin. Like Jeff, Rand was set to share the interesting search ranking factors findings from a study recently conducted by SEOmoz. There was two types of data coming at us: 1) opinion data, straight from an SEOs survey, and 2) correlation data— the way the study panned out.
Correlation Data Methodology
In 2009, 65% of surveyed SEOs thought link factors were amazingly important. In 2011, it’s more like 45%. Serious drop off.
What do SEOs believe will happen with Google’s use of ranking features in the future?
- Exact keyword match will decrease
- Anchor text in external links will decrease
- Prominence of ads vs. content will increase
- Usage data will increase
- Social signals at a domain level will increase
- Perceived value will increase
- Social signals at page level will increase
What the study shows:
- Exact match keyword domains are losing importance as search ranking factors
- Google is getting better at discounting the value of paid links
- Domain level link data is surprisingly similar to page level link data in correlation
Have exact match domains lost their luster?
- Yes, to a degree– exact match domains have fallen considerably in the past 10 months
Is Google evil?
- Google has said that linking externally is good, slow pages are bad, using Google services won’t give any benefit (eg: Google Analytics)
- SEOmoz’ study & data supports those statements
- Study finds by-and large, there’s not much “evil” in Google’s rankings, none that correlation research reveals
Now, onto social signals – from users of Twitter to Facebook & Google Buzz. Get ready for your mind to be blown.
- The # of FB shares is the single highest correlated metric with higher Google rankings
- After FB shares, it’s the sum of FB shares, likes, & comments-– after that, the # of linking c-blocks to pages
- Rand was amazed that FB share data was present for 61% of pages in the top 30 results
“Stuff gets shared on Facebook a crapton. That’s a technical term.” – Rand Fishkin
- Opinion data suggested that FB shares are highly predictive of link activity, but correlated data didn’t exactly echo that concept.
- Regardless, as Rand said, “This social thing has got some legs.”
- Takeaway: Twitter’s correlation wanes, but FB features appear influential. FB deserves more SEO attention than it currently receives.
And lastly, a special treat for you – SEOmoz’s ranking factors raw data available here –> http://bit.ly/kzDt7J
Next up to the mic, Matthew Brown, set to discuss the new SEO elements: brand, social, & local.
Word of caution: just because these elements are “new to the mix,” don’t underestimate the tried and true elements:
- Links are still the engine that will drive your SEO.
- On-site optimization is the gas that fuels it.
- These things are still vital, but new things are getting added, and blended together with the entire mix of elements.
So, once again – brand, social, & local. Maybe they’ll help you rank better, maybe not. Regardless, they’re shaping the SERPs, and as such, worth paying mind to.
Brand is a compound of link quality, link text, social shares, reputation, authority, history, country, locality, etc.
- Type in “microwave” in Google… not even “buy microwave,” just “microwave,” and basically all your SERPs are relating to transactional links.
- Ranked #6 on the SERP is Panasonic’s product page for microwaves. It’s hideous. Horrible SEO.
- Why does it rank? Panasonic is popular brand, has authority, trusted.
How do engines find brand signals?
- Searches for brand or branded KWs in toolbar or address bar
- Brands are going to have a lot of exact and partial match anchor text for their names and variations
- Building social signals is a great way to build brand signals. How to do this?
- Establish / enhance your Facebook presence, for example.
BUT… how to do build brand and social signals on MY site? Facebook marketing campaigns are great, but how do you bring that back to your own site? You deserve the equity!
Check out ThinkGeek.com as a noteworthy example. They’re bringing sexy back social equity back by leveraging the Open Graph, i.e.: object based social signals, i.e.: they have the Like button directly on their page. All those likes their content gets STAY on the page, helping search engines recognize social signals on their OWN(ed) site.
- Remember, Bing & Facebook are BFFs. Ergo, use the Open Graph!
Also: identify your site’s social authority hubs… which social networks work for the types of content you share / site you own?
- Are there lots more of LinkedIn shares vs. FB shares vs. tweets for your content? Make strategic decisions based on this data.
- How do you want to encourage people to utilize social networks? Take note of WHERE they are – invest your time there.
Tip: Schema.org – check out this baby, stay current on semantic goodies.
Last but not least – localization. Matt asks, is localization the online brand killer? It does seem to dominate the SERPs… sometimes even pushing Wikipedia down wayyyy below the fold (something we all secretly wish for, right? ehh…)
- Google Suggest is one way to understand if your keywords are becoming increasingly localized
- Mergewords lets you mash up cities (any location, really) with KWs – run those through Google to glean inventory / increase / decrease, etc.
First mover status: Take chances, make guesses, the ones that pay off will be worth it! Example, Google +1 doesn’t seem like much now, but you don’t want to WAIT for it to become super important – start now!
And that’s all she wrote! (Or he… this was a rather male-tastic panel.) Stay tuned for more blog coverage coming atcha from #SMX Advanced, and live-coverage via @beebow.