A Data Science Central Community
#1 Online reputation protection
Helping companies find bad comments posted by disgruntled employees or competitors, and fight back by diluting the bad content, via automated posting of positive messages in relevant blogs. Also helping individuals recover from postings (from themselves or other people) that might harm their job prospects.
#2 Social network competitive intelligence
Traditional measurement companies (Alexa, Quantcast, Compete, Comscore) use old-fashioned and inconsistent metrics that do not capture the quality of the traffic being measured (e.g they do not filter out all robots or do not discriminate between spam and good content).
New model: use a web crawler to scrape millions of blogs or publisher websites, and measure popularity using metrics such as number of advertisers and list/quality of advertisers, number of users who did not post anything this month but posted something over the last 6 months (an indicator of churn), number of users who posted this month but not over the last 6 months (an indicator of new member growth), number of users posting on a regular basis (an indicator of retention / faithfulness / quality), average length of posts, number of replies per post, average number of rare keywords per post (an indicator of level of specialization), etc. Provide this service as competitive intelligence to competing publishers and/or advertisers.
You can also use the same web crawler to tell businesses what people are saying about their products.