Is Google preparing for an online reputation service?

You might remember the fuss that started at the end of last year due to the practices of the DecorMyEyes company. In an article in The New York Times, the Russian owner of this company had explained how he scored very well in Google's search results, due to complaints and negative online comments that were intently and systematically provoked.

Danny Sullivan of Search Engine Land said at that time that Google could prevent such practices by incorporating the results of Google Product Search in its general search results. That way, sites that receive a high number of negative comments and complaints could not profit any longer from a good positioning in the search engine.

This brings us in the field of sentiment analysis, something that is primarily seen in social media monitoring.

Google could (and would) punish websites that get lots of negative comments and thereby stop this ‘anything is good to get lots of visitors’ methods. However, as Brygier writes, if it were to do this, also sites with, for example, controversial or political content would be punished. So the company decided to blacklist a limited number of companies that receive lots of online critics.

Last month, Google announced a new function: ‘reading level analysis’, whereby the company can determine a website’s reading level. In the increasing use of text analysis that goes together with this, there also is a high amount of sentiment analysis. Moreover, the company disposes of several different techniques to incorporate social media parameters in its different services.

Brygier gives an interesting example of how Google could use all these technologies in order to, for example, determine the influence and the sentiment of content in Google Alerts.

So he sees the company evolve further in the direction of social media monitoring, be it in partnership with an existing media player, or by themselves. For both scenarios, he cites intelligent arguments that can be read here.

Google in reputation and sentiment analysis and in social media monitoring. It definitely is not a silly idea. What do you think?