Twitter: when will we get TweetRank?

Google has PageRank to tell us what web pages are the best. Facebook has EdgeRank to show us the most interesting, relevant social updates. But what does Twitter have?

There are a lot of tweets.
That might be an understatement. Let’s put it this way: THERE ARE A LOT OF TWEETS. And the signal to noise ratio is not as good as it could be … nor is it as good as it should be.

  • in February 2011, there were 155 million tweets per day
  • in March 2012, there are over 500 million Twitter accounts
  • and they send out 33 billion tweets a day

May I repeat myself … that’s a lot of tweets. WAY too many for anyone to read, and hard even to try to imagine.

That said, many tweets, if not most, suck
We’ve all seen the account from the “social media expert” which consists of 100K followers, 100K following, and mostly automated tweets from a quote database … over and over and over. Profoundly uncool, and largely uninformative, not to mention unimaginative. In a word, spam.

Then there are others who add nothing, but retweet everything. They are slightly less annoying, but they are adding more noise than signal.

Most egregious are the twerps, er, tweeps who endlessly promote their own content, much of it consisting of crappy digital stuff to buy. No conversation, no engagement, no community just a carnival barker set on perpetual play.

These people kill the Twitterverse.

There’s gotta be something better
There’s gotta be a way to increase the signal to noise ratio. To dampen the static, and crank up the song … let the good stuff win out, rather than being drowned out.

There’s a bunch of options for this already. None of them are ideal.

One of them is simple: follow fewer people. If you only follow a few, you’re guaranteed to get less noise, right?

Right, but you’re also guaranteed to get less signal, too.

Even if you only follow a few top-notch people in whatever niches your interest lie … you’re still going to get:

  • @ messages from people you follow to other people who follow them … all of them of dubious importance to you
  • some degree of lame and inconsequential updates (we all do it … we’re HERE, this steak is AWESOME, my toes hurt, etc. etc.)
  • irrelevant messages … just because you follow an interesting person doesn’t mean that ALL of his or her interests directly overlap with yours

It’s true we should follow smart, but Twitter should show smart, too
There are cues to how important or relevant a tweet is. Some are obvious, such as these public indicators:

  • social importance of the tweeter
  • number of retweets or mentions
  • relevance to similar info in the Twitterverse

Some are not so obvious. They’re personal, related to complexities of how your interest graph and the person you are following intersect. So, for instance, if Twitter noticed over time that you responded more to Scoble’s tweets about open source, rather than Microsoft. Or more about startups than gadgets.

In both of those ways (macro and micro, public and personal), Twitter could shape the tweets you see to be higher quality, more relevant, better. There is a danger that we do get trapped in our own little reality bubbles, as Eli Pariser has warned. But the TweetRank algorithm could introduce some serendipity, perhaps bringing in high-quality tweets from other’s interest graphs, rocking your world with new ideas and perspectives 🙂

It just seems odd that Google has a TweetRank system but Twitter does not. And the Twitter user experience could be improved with a gently, carefully applied algorithm to improve the overall quality of tweets people see.