Conversation Analytics applied to Twitter

Tweeterboard_logo Gene Smith, a principle at nForm, has developed TweeterBoard by calculating a conversation "leaderboard" of twitter users. 

As I’ve outlined here a number of times, there are significant new opportunities to create value for users at the intersection of conversation and analytics.   The technologies have developed, the markets are migrating toward a data-analytics model and users are ready.  The next 3 years will see the emergence of new conversation analytics applications for consumer, prosumer and enterprise markets.  We’re right on the cusp. 

Think about it this way.  Who we communicate with represents our real social network.  It is this network of people with whom we converse that holds the social value.  These people warrant our attention and we engage them in on-going discussions, dialogs and conversation threads on a daily basis. It is this flow of phone calls, IMs, emails and tweets that represents our real social network.  Nodes and Edges… People and Conversations….are the real sources of value as Enterprise 2.0 emerges.

I like Tweeterboard because it is a further proofpoint of the end-user value found in conversation analytics.

Marshall Kirkpatrick of RW/W summarizes the approach Twitterboard takes to calculating the leaderboard:

"Your influence is determined by the influence of people who talk about you publicly, apparently, though it’s not completely clear.  Tweeterboard tracks who replies publicly to a given user using the @ symbol and their user name, and who they reply to as well. It then uses a secret algorithm to assign a number of points. Generally speaking, it appears to work like the premise of Google – value is conferred to you through links from people who have had high value conferred to them by links from people who have… "   

In response to the write-up by Marshall, Gene addresses some questions in the comment section:

"Hi Marshall,

Thanks for mentioning Tweeterboard. You’re right that it looks at @ replies and, for now, only tracks public updates.

Generally speaking, it appears to work like the premise of Google

Yes, it’s very similar to Google in that your reputation is calculated based on who talks (links) to you.

The leaderboard is changing a lot right now though; presumably time and algorithm adjustments will slow that change down a little bit at least.

Indeed. This is the third version of the reputation calculation, and it’s more stable than the earlier versions but still pretty volatile. I have plans to make it more stable… hopefully that will happen early in the new year.

Steve Rhodes: I’d be careful about calling them the 100 most influential people on twitter particularly since we don’t know the method the site uses to calculate the rankings.

A good point. Tweeterboard gives you just one of many possible measures of influence. But the basic idea is that conversations–who talks to you–is a better measure of influence and engagement than followers, friends, and so on. Ideally, people like Andy Carvin wouldn’t be able to sustain a high reputation unless they were really engaging their followers.

Posted by: Gene | December 19, 2007 4:32 AM "


  1. Good post, thanks. I like the idea of conversation analytics.
    Social sofwtare opens the eyes as to what can be achieved, but it has left us with fractured interfaces to those conversations.
    Compared to the new tools, CRM is weak, web analytics aren’t mainstream and email just doesn’t cut it for anything other than one to one.
    It’s almost like magnetic fields are changing, and we’re watching the compass

  2. Matt,
    ….”magnetic fields are changing, and we’re watching the compass”.
    That’s an interesting way to think of it. While I may not agree 100% with your vision of the PBX future , I do appreciate your perpectives on Unified Communications outlined in your blog. Keep the great posts coming and thanks for your follow-up.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s