10 comments

• 

  craig donato

  3 years ago

  i think what edgeio is doing is interesting. if you agree, i'd also encourage you to look at(and comment on) a recently proposed microformat specification called hlisting http://www.microformats.org/wiki/hlisting-proposal. This would enable a wide variety of engines (Oodle, Technorati, Google) to easily find and reference a classifieds listing posted on a blog.
  
  You might also want to check out an alpha version of an open-source plug-in for WordPress that uses this spec -- http://blog.labnotes.org/2006/02/07/posting-a-l...
  
  interesting times...
  reply
• 

  Mathew

  3 years ago

  Thanks, Craig. I will definitely do that.
  reply
• 

  Vaibhav Domkundwar - iNods.com

  3 years ago

  Mathew:
  
  My company iNods is in a similar edge content aggregation and search space. Edge content aggregation can really delivery a lot of value but there will be a lot of challenges in terms of consolidation and discovery of that edge content. Infact this is the real challenge for more edge aggregation plays.
  
  The idea of asking users to tag their classifieds with "listing" tag is good but its not trivial to achieve on a large scale. Secondly, if the whole idea depends on this then its tough to make it work. There is biggg chicken and egg problem here. Now assume that people start tagging their classfieds with "listing" tag - in this case it will be trivial for blog search engines to roll our a similar service in a fraction of the the time and with a much wider coverage.
  
  In almost all edge plays, I believe it is critical to have a stronger barrier to entry and/or some other community angle that makes it unique. We do have a similar challenge at iNods (we think we are solving the edge content discovery problem is a more defensible way) and you will see how we are addressing it with a range of our upcoming features.
  reply
• 

  Mathew

  3 years ago

  Thanks, Vaibhav. iNods sounds like an interesting experiment.
  
  Mathew
  reply
• 

  Markus

  3 years ago

  I think there are some serious serious flaws with this.
  
  Like craig I block several hundred con artists and scammers per day. Ebay, Amazon etc all use IP address and X_forwarded_for etc to detect 95% of the cons and spam. Once you go to a distributed system you lose the ability to detect most of that stuff.
  
  The nigerians and russians etc are going to create fake sites, and flood edegio with listings. You will then be lured to those other sites and enter your CC etc. Think Ebay email phishing turning into blog listing phishing. Ebay employs over 1,000 people to stop scams. A site like edegio would need at least 10 to 100 times the work force to achieve the same effect.
  
  All i'm really saying is this site would have a lot of growing pains, and a lot of bad publicity. Now assuming it does well what exactly is stopping technorati from assigning 10 developers for a week and clone the system?
  reply
• 

  Mathew

  3 years ago

  Good points, Markus.
  reply
• 

  John Furrier

  3 years ago

  EdgeIO main founder besides Mike is Keith Teare. You article makes it sound like Mike's startup when actually it's been Keith Teare and Mike Arrington who put it together.
  
  EdgeIO is certainly poised to be disruptive and the market is ripe for it
  reply
• 

  Markus

  3 years ago

  Thanks Mathew,
  
  I should add some more context to my comments. I run one of the top 10 sites in canada, if not the largest. Plentyoffish.com which has over 13 million pageviews/day. I am also the largest 1 person internet company and have been extremely disruptive basically destroyed the paid dating model in Canada and now starting in the US.
  
  1. Existing players will view this company as Free R&D
  2. They will have several clones within weeks if it gains traction.
  3. The scam problem etc may seem small now. Trust me if they gain any sort of traction they are looking at 90% scams. Since they are collecting listings and people selling stuff the abily to detect scams is extremely low. How can you really detect scams from collecting a pile of spam ?
  
  4. Existing scam detections such as used by all major retailers can not be used.
  
  I believe if they want to be truely disruptive they should go out and collect peoples resumes and build a metabase of those. There is enough critical mass there that it would take off FAST, given that something like 30% of bloggers are doing it to advance their career.
  reply
• 

  Mathew

  3 years ago

  Thanks, John -- you're right, I should mention Keith too.
  
  And Markus, thanks for the additional info. I added an update to my original post noting that the task Edgeio has set itself is not going to be easy. I wonder if they will take you up on your resume idea.
  reply
• 

  joe

  3 years ago

  Markus brings up an interesting point....when postings are supposed to be ads, how do you distinguish between legitimate ones and spam? Is there a meaningful distinction?
  
  Flat-out fraud would count as spam, of course. If there's some kind of classification system (electronics, books, etc) then purposely mis-classified items would be spam. Beyond that, I don't know.
  reply