Robert Scoble's applecart

'Google doesn't understand social behavior' is the premise upon which Robert Scoble developed his ideas about how Mahalo, Facebook and Techmeme will trounce Google over time. I'm no more convinced than most of those who are cited on Techmeme except in one area I regard as a green field - business.

Robert and I discussed this in a Skype conversation. At the time, comments were flowing past the kyte.tv videos at breakneck speed and it was almost impossible to keep up. Robert acknowledged it was incredibly difficult to keep up with the flow.

My big sticking points are scale and selection. Robert acknowledges that people don't scale well. That has particular relevance for Mahalo and Techmeme but would be of no concern to knowledge based businesses where shared information is the basis for differentiated service and quality. On selection, Techmeme has a list of 'favored' sources. While that's not a bad thing, I always wonder the extent to which it misses important material. If you have a narrow focus then Techmeme probably misses a lot of the things that are important to you. It misses many quality debates around enterprise applications. In a business context, that would not be an issue because there is no reason why a directory of all contributors/contributions could not be maintained, indexed and related.

During our conversation, I bemoaned the fact that so many services - and Facebook in particular - seem hell bent on addressing consumers while eschewing business markets. I can understand why. Making a service massively viral is not likely in the business sector. Conventional wisdom suggests that new entrants need huge buzz which usually only happens in consumer markets. When I look at the truly business useful Facebook applications, like Google Shared Items, it has a very small number of takers. Yet that could be the genesis of a Techmeme for business.

Business is different. It wants information that can be turned into knowledge. Past attempts at knowledge management have not fulfilled that promise. If Mahalo, Techmeme and Facebook were to turn their attention in that direction, I'm betting they could make some serious early bucks. Robert said he gets the sense Facebook is looking at this segment but could not provide any additional meat. Only time will tell.

In the meantime Dan Farber discussed the efforts of companies that barely register in the cacophony of those commenting on Techmeme but which are doing important work in this area:

The key is not to expect Google, Yahoo, Microsoft or some as yet unknown company to have their blinders on and cease to improve their search engines. Powerset, for example, is applying crawling, linguistic analysis with natural language processing, semantic search and human engineered wisdom of the crowds to build a search engine that can deliver more accurate results. Hakia is also taking a semantic approach to searching the Web.

I'm not convinced that Google in particular is doing as much as Dan implies. As I understand it, most of Google's development resources are concentrated on fleshing out the ad engines. In those companies where I have seen Google appliances, there's not a huge amount of enthusiasm for delivered results.

As information stores expand, the need for reliable search, melded to social networks takes on increasing importance. It should not therefore be surprising to see a comprehensive list of players mentioned in the upcoming Enterprise Search Summit agenda. Some like Microsoft, Oracle, OpenText and SAP will be familiar. Others like Bainsight, Convera and Coveo are less well known.

While I see difficulties in the wider world for the model Robert suggests, he is not clueless. His thinking may be formative and he's maybe not looking in the right direction for my taste but as Dan says, he's started a useful conversation.