AnalyticBridge
A Data Science Central Community
Social networks: is the growth real? A business opportunity.
According to several traffic measurement companies (Quantcast, etc.), Twitter's traffic is exploding, followed by LinkedIn, Facebook and other networks. Once you factor in and remove spammers, growth numbers look much more modest.
From our experience, we see very strong growth on Twitter, but after removing spammers, the growth is close to 0. LinkedIn shows a strong growth too, smaller than Twitter, but after removing spammers, LinkedIn growth appears much stronger than Twitter.
Facebook is not working that well with us - growth is very modest because Facebook members are not interested in business. However, we have found very little spam on Facebook, compared with Twitter and LinkedIn. For us, LinkedIn is definitely the winner, after factoring in spam and automated traffic that is not accounted for by Internet traffic measuring companies. Ning, which provides strong growth to AnalyticBridge, also brings its share of spammers - more than LinkedIn, but much less than Twitter.
I believe that eventually Twitter will clean its traffic. But for now, there is a great opportunity for start up entrepreneurs: creating a tool that allows Twitter, LinkedIn, Facebook and Ning network creators to automatically flag members that have a bad spam score, and allows administrator to automatically block / remove / ban members with very high spam scores.
From our experience, we see very strong growth on Twitter, but after removing spammers, the growth is close to 0. LinkedIn shows a strong growth too, smaller than Twitter, but after removing spammers, LinkedIn growth appears much stronger than Twitter.
Facebook is not working that well with us - growth is very modest because Facebook members are not interested in business. However, we have found very little spam on Facebook, compared with Twitter and LinkedIn. For us, LinkedIn is definitely the winner, after factoring in spam and automated traffic that is not accounted for by Internet traffic measuring companies. Ning, which provides strong growth to AnalyticBridge, also brings its share of spammers - more than LinkedIn, but much less than Twitter.
I believe that eventually Twitter will clean its traffic. But for now, there is a great opportunity for start up entrepreneurs: creating a tool that allows Twitter, LinkedIn, Facebook and Ning network creators to automatically flag members that have a bad spam score, and allows administrator to automatically block / remove / ban members with very high spam scores.
Comment
-
Comment by Vincent Granville on November 28, 2009 at 9:59pm -
It's based of our experience on Ning (6,400 members), LinkedIn (> 10,000 members) , Facebook (900) and Twitter (900).
On Twitter, people are using bots to automatically follow other people -- anyone. I have to block these followers (more than 70% are true spammers selling software to grow your list of followers, or selling get rich quick schemes), otherwise, my profile would get diluted to the point of becoming worthless.
LinkedIn attracts lots of recruiters, but as long as they post analytic jobs in the job section, and respect some basic etiquette (being professionnal), that's fine with me.
-
Comment by Theodore Omtzigt on November 28, 2009 at 12:08pm -
Very interesting data: can you elaborate on how data is collected and analyzed?
The Twitter spamming problem is very real as anyone that has a Twitter account can attest to. Over on another discussion we talked about the business model to generate revenue on these social networks is likely going to disrupt their basic use cases and thus potentially create a backlash.
Given the impact that spam has on the value proposition of these services, the core web properties will need to clean their own shop. Third party technology will have a hard time to impact internal politics. Secondly, good technology demonstrations will be difficult to orchestrate since most of the data is not accessible. And finally, the piece that we constantly run into, with spam filtering you really want to have this occur in real-time, and typically, the computes are not available to do this in the critical path.
© 2019 AnalyticBridge.com is a subsidiary and dedicated channel of Data Science Central LLC
Powered by
Badges | Report an Issue | Privacy Policy | Terms of Service
You need to be a member of AnalyticBridge to add comments!
Join AnalyticBridge