SMX Advanced, the “Give it Up” Session (part 1)

Thursday, July 5, 2007
Posted by Jim Hedger @ 11:11 am

The final session at SMX Advanced Seattle came with an information embargo attached. Attendees were asked to respect a one-month silence period, to refrain from blogging information from the session. Though the conference-space was packed with a standing room only crowd, the embargo has been universally observed. Today marks the one month point. The embargo has expired and “Give it Up” posts are starting to appear on SEO/SEM blogs.

This was supposed to be the session at which leading edge SEOs released their deepest and most effective search marketing secrets. Rumour had it the major search engines were to be banished from the room. Any blogger known to post info from the session before the embargo ended would be considered persona morte by others in the industry. No kidding around.

The speakers list, which was not posted to the SMX website, featured; Stephan Spencer, Sheri Thurow, Mikkel deMib Svendsen, Bruce Clay, Mike Grehan, Jill Whalen, Jennifer (JenSense) Slegg, Todd Friesen, Greg Boser, Danny Sullivan, and Matt Cutts (yes… that Matt Cutts).

Danny warmed the crowd by passing an old story about beating the submission wait-times at Yahoo. Since so many webmasters were submitting their URLs to Yahoo!, a priority queue was established. The secret to passing to the head of the line was to write (yahoo) after your email address during submission. While nobody openly discussed this trick, it was known by a number of SEOs, spread through the community by word of mouth.

This is where he warned eager writers to keep their fingers still for 30-days. “On your honour people, don’t blog this. Keep it under wraps for one month from now; Greenwich Mean Time. If anybody breaks the embargo, they are dead to us.”

While the rumour said search engine reps would be banned from attending the session, several were in the room. Matt Cutts pointed out that the “room” had cloth walls and that he could easily stand outside and still hear. In exchange for a quick story, he (and other SE reps) was allowed to stay.

Matt’s story involved a spammer from Eastern Europe who found a typo-exploit and got his hands on some 7000 domains with the PageRank of eTrade, for free. It sort of goes like this:

When you buy a domain name, you need to put it somewhere. Since many domain-owners are buying before using, their domains are often registered to the nameserver “lamedomain.org”, a holding pen for otherwise unregistered (or unhosted) domains. Now, as every webmaster knows, typos happen and, as it happens, thousands of domains were set to the nameserver “lame-domain.org”. Since this is a spam-story, you can guess who registered the domain Lame-Domain.org; the spammer from Eastern Europe. Suddenly, he controlled seven thousand domains which he did not own and which all had high PageRank values. Needless to say, since Matt was sharing this story, the exploit had been caught and shut down. The audience decided to let him stay anyway. :)

Next up was Stephan Spencer who had the most to say. Stephan talked about:

How to eliminate competitors’ grouped results by determining the actual indexed position of documents from the same domain when they appear grouped together on SERPs pages.

How to use Google translate (English to English) to view the true content of a potentially cloaked page.

How to eliminate all supplemental results from SERPs by changing search preferences to English only.

How to obtain free analyst reports from Forrester Research by going to Google and typing, “forrester research grapevine endnotes filetype:pdf (or ppt)

How to use sites that are one click away from Google to get stronger link love. Look for high PR sites that give links to sponsors or patrons. NGOs or even better, Google philanthropic partners are suggested.

How to use search friendly cloaking techniques that strip userID or parameter identifiers off URLs. Stephan specifically says to avoid content cloaking in which different sets of content are served to spiders than to humans.

How to keep your current link network updated and to spot opportunities within it to refresh anchor text.

How to display all supplemental results for a specific domain by searching for, [www.YOURDOMAIN.com*** -qerdafjdfs].

Sheri Thurow came up next and made three suggestions.
Her first was Education. Sheri shared how taking a course, Methods of organizing information for retrieval and display, altered her perspectives on search and search marketing.

Her second was Read Articles. Two she specifically recommended are:

The Design of Browsing and Berrypicking Techniques for the Online Search Interface, by Marcia J. Bates, 1989

What is a “document”? by Michael K. Buckland, 1997

Sheri also urged her audience to move beyond the “cat-and-mouse mentality” of algorithm chasing and trick-based SEO saying advanced searchers need to evolve.

Next up, Mikkel took the stage threatening to give out Matt’s phone number. Fortunately for Matt, it was only a threat. Mikkel’s message, however, was very real. To paraphrase it as best I can,

Search engine placement relies on content generated to please search engines. There is absolutely no way to naturally create the volume of content necessary for placement of documents in extremely large sites. It isn’t humanly possible and even if it was, there is no way a human can be expected to perform the necessary tweaks to each individual piece of content. The solution… Computer generated content.

Mikkel went on to speak of using Markov chain mathematics to create computer generated documents best suited for their keyword phrases on the fly. To sum it up, the Markov chain is used to examine and predict probabilities of change in a non-static property. In the instance of creating search-ready documents, it goes like this:

Scanning scraped contextual text, you can see and predict incidents of words or phrases appearing against each other. “Sit in” could relate to the word “chair” ten times and “college” twice. When examining other documents using the words “sit in”, Mikkel’s technique enables him to electronically generate contextually relevant documents. Mike Grehan noted that the concept of PageRank is also based on Markov chains.

Next on stage was Bruce Clay who spoke of internal linking silos, forcing SEO training on clients, the future of search marketing and experimentation with traffic generation techniques.

Bruce was followed by Mike Grehan who’s basic message was,

Search has expanded beyond text on page. Think different. Think about images, videos and why users click on specific search results.

Mike also stressed the importance of the value of links. Quality is going to be more important than quality. Look for links from traditionally off-line sources.

Tomorrow, notes from Jill Whalen, Jennifer Slegg, Todd Friesen and Greg Boser’s presentations.

No Comments »

No comments yet.

RSS feed for comments on this post. TrackBack URI

Leave a comment