Tag Archives: stats

An algo for Twitter Authority

This post from Loic Le Meur has sparked a lot of debate that Twitter should have a “search by authority” feature. Keeping the ego factors involved aside “search by authority” could indeed be useful for finding out what people who matter are saying about a particular topic. This will be particularly useful when finding what thought leaders are saying about a particular thing/topic and getting selected and relevant information during crisis times amongst others.

As mentioned here, follower count isn’t the right metric to gauge authority and while twitter ratio(followers/following) could be a better indicator of a one’s celebrity status doesn’t imply authority. Though there is almost zero possibility of Twitter implementing this but just for fun lets see what could possibly be a nice algo to determine one’s authority on Twitter.

1) Re-Tweet Ratio(RTR): RTR = Total number of re-tweets/Total number of tweets. A high Re-Tweet ratio can imply great content, large following and thus high authority. Going a step further if instead of posting something original, you re-tweeted someone’s content which one of your followers further re-tweeted then its credit should be added to the original posters authority and not the middle man i.e your authority will ony include re-tweets for your original content.

2) Tweet Favorite Ratio(TFR): TFR = Total number of favorites one’s tweets got/Total number of tweets. A high Tweet Favorite Ratio is another indicator of great content, large following and thus high authority. Since unlike kwippy where favoriting(and commenting) stats are out in open

, twitter isn’t open about favoriting information and you not even know if someone liked your tweet or favorited it.

3) Tweet Reply Ratio(TRR): TRR = Total replies one got/Total number of tweets . A high Tweet Reply Ratio might not be that clear an indicator about great content or large following it definitely indicates a high level of engagement which in turn can loosely be linked to one’s authority. For ex: this tweet by Chris Brogan
sent two days back got hundreds of replies and still continues to get replies.

4) Tweet Link Backs Ratio(TLBR): The most far fetched and away from reality measure could be TLBR. TLBR = Total number of link backs one’s tweets got/Total number of tweets. Though a link backs for one’s tweets aren’t tracked, if done we could include this data to calculate TLBR. A lot many times people’s tweet inspire blog posts and discussions on various aggregators. A high TLBR can also serve as a indicator about one’s authority.

Now we have atleast four parameters which we can use to calculate one’s twitter authority. If there’s a formula for calculating twitter authority it might look like this

Twitter Authority = 2 * RTR + TFR + TLBR + 1/2*TRT

Since Twitter Re-Tweet ratio is the highest measure of great excellent it’s given most weightage. Twitter Favorite Ratio comes next and is almost same as relevant as Twitter Link Back Ratio they have equal weightage. Twitter Reply Ratio, doesn’t imply great content necessarily but a tightly knit follower base it gets the least weightage.

Its was a fun post written just like that and shouldn’t be taken too seriously but if you did, tell me what you think about it ?