An Algorithm for a Social Fabric

I just got done watching Robert Scoble’s talk on Social Graph Based Search. While I find it very compelling, there are a few points that I take issue with. Before I get into that, I want to first note the overwhelming sense of “eventness” or Sobytiinost I felt while when watching this video. The influx of information occurring on Kyte.tv adds a great deal to the conversation. I would love to hear what Brad Berens has to say about the Kyte.tv experience.


Issue’s I have with Robert Scoble’s point of view…




  •  Each reference to SEO has negative connotations.




oWhile I understand there are some bad practices out there, there are many SEO’s who are marketers trying to disseminate their information.


o Not every web designer is good at coding pages for search, and not everyone has as many friends as Robert does (at least in the social media space. Remember, we are still in the early adopter phase and many of us are still talking in an echo chamber) and it is more difficult to get them up the social graph (at least in the initial phase of marketing).


oRobert, I totally see your point, but SEO is not all noise.




  •  I do believe that social fabrics can lead us to more relevant results, but…




o Robert, you make a point of how you have 5000 friends and Walt Mossberg may have 100,000. The cause of this is Mossberg benefits from the exposure he gets from the Wall Street Journal, a major media company. This exposure is synonymous to a monster inbound search link (and a paid one at that).


o There will always be power structures in the media.




  • I am not saying I am for or against them, but as power structures form, a certain level or corruption is inevitable (at least this is what history teaches us).




  • Once a person reaches a certain level of power it is very difficult for the masses to overthrow them, for the masses trying to climb the ladder are too busy making nice to these figures and working their way up.




  •  I can point to examples of this in the new social media space (although I don’t care to do so at this time)




Robert, thanks for this talk!


Whether I agree with all, some or none, it really got me thinking, and I appreciate this above anything else.






MY NOTES ON SCOBLE’S TALK (not editorialized)



Social Graph Based Search (and why it will up end the search system in four years)




  • Who are Google’s Main Competitors


    • Techmeme-  Creation of “the fabric”

      • Vertical search based on a limited, hand selected number of sources


      • SEO resistance

        • Either in the “Techmeme” or trusted by the Techmeme pool

        • The chain of trust—hard to break into






    • Mahalo-

      • Human beings are better at displaying stuff than machines

      • Offensive aimed directly at Google




    • Facebook- Why does it matter?

      • Trust Level








  • The Problem With Google




  • Google works off of <Title><H1> <P1> and inbound (of course there are hundreds of other factors, but these are the ones that Scoble mentions) links that are, or are not relevant (link baiting etc.)




    • Scoble refers to Google as “noise”




      • Talks of the trouble of betting around the noise of Google






    •  Refers to SEO as noise






  •  “Google does not understand social behaviors




    • Says that they cannot due to business related factors








  • What Can Be Done?


    • Mahalo should be building on top of Google’s algorithmic approach

      • Understand the old school




    • What if Mahalo hooked up with Facebook to build a fabric?

      • Building a fabric of affinity




    • We need and open social graph

      • Facebook is a walled garden and this is a problem



    • Facebook has figured out how to lock out SEO’s

    • Techmeme is building a pure fabric due to lack of SEO’s






  • Why Four Years?

    • It is going to take two years for Mahalo to be recognized as a top search engine (same with Facebook, Techmeme)

    • It will take another two years for people to catch up






  • The problem of scaling


    • The community and the exponential number of those related to the community (the output decreases at each exponential levels)

      • Does not seem to scale


      • What if everyone were to have there one little Mahalo


        • Facebook

          • An algorithm for a social fabric











Tags: , , , , , , , , , , , ,  , , , , , ,