Twitter’s Engagement Problem [Data]

Twitter Engagement - Are you talking to me?Twitter has an engagement problem.
A microblogging social media platform enabling 140 character communications, Twitter is approaching its 500 millionth account. That makes this is a good point to look at what new Twitter users are actually doing.

A top line view of Twitter’s path to 400 million users reveals slow adoption and then a big jump in early 2009. This makes sense since, unlike other social media platforms, Twitter has its own special language and requires time to become proficient. [Need help? Here’s how to get your Twitter Mojo.] As Twitter’s gained more mainstream attention, it’s continued on a relatively steady growth trend.

Interestingly, during these last two years, teen use of Twitter has doubled from 8% of online teens to 16% according to Pew Internet Research. This teen segment has moved to Twitter since it gives them a sense of control, allows users to select a pseudonym, and offers a channel for private communications. 

Looking at the behavior of recent Twitter registrants, issues related to engagement emerge. Analysis of 100,000 newly registered accounts, starting from account number 400 million, by Twopcharts revealed the following trends.

  1. 12% of these accounts were deleted, cancelled or suspended leaving 88,052 remaining accounts. By itself, this number shows that actions are being taken to eliminate spam and bots.
  2. 62.3% of these remaining accounts are still eggheads because they haven’t changed their profile image. Contrary to accepted social media practice where transparency is expected, these accounts are anonymous to other Twitter users. Across the active accounts, 20.2% have filled in the location field and 17.3% have filled in the bio field.
  3. 4.4% of these remaining accounts hide their tweets from the public and have protected their accounts. There’s a good chance that these accounts are being used for private direct messaging and/or monitoring specific profiles.
  4. 24.3% of these remaining accounts don’t follow anyone else. Since most social media participants use these platforms to engage and communicate with others, this lack of activity is surprising. That said, they may use a third party tool to track people, hashtags or keywords thereby gleaning the information they desire.
  5. 52.7% of these remaining accounts have no followers. This means that these members aren’t engaged with a community of other like-minded individuals or even have any private conversations. They can extend their tweets through the use of hashtags but, without a following, it’s unlikely anyone will pay much attention to them.
  6. 53.8% of the sample never tweeted. In a social media ecosystem filled with conversation and noise, it’s hard to believe that this many Twitter participants have nothing to say to anyone. It’s more likely that they’re challenged to overcome the hurdles Twitter participation presents.
  7. 12.3% of the sample sent one tweet. Over 80% of these Twitter registrants sent their lone tweet over two months ago. Therefore it’s a good bet that they aren’t about to return. This makes sense since Twitter has its own language that makes it difficult to jump in and come up to speed. It reminds me of when my mother first got email and couldn’t understand why to use it when she could make a phone call much easier and quicker. Unfortunately for Twitter this behavior isn’t new. [Check out Nielsen’s analysis of Twitter Quitters from 2009]

Twopcharts estimates that only a maximum of 20%-25% of new registered Twitter users convert to active new users, with only about 10% of new accounts actively sending out tweets. Before you assume that all Twitter activity is created equal, Twopcharts explains that these accounts were opened in the last three months. While they’re probably similar to accounts opened within the last year, their activity can’t simply be extrapolated to the entire Twitter population.

What does this mean for marketers? Take this low usage into consideration when developing your social media marketing plans. As with any emerging technology, understand that you may need to provide an explanation to your target audience if you want them to overcome the hurdles to using Twitter.

How do you view these results as they apply to your marketing?

Happy marketing,
Heidi Cohen

Here are some related articles you may be interested in.

Photo credit:
Tags , , , , , . Bookmark the permalink.

6 Responses to Twitter’s Engagement Problem [Data]

  1. Dan Farfan says:

    I don’t know how much of the analysis text was taken from the twopcharts report linked, but most of it is flawed, imo. The primary mistake is it all assumes that people, individual humans, hold the accounts that were not “deleted, cancelled or suspended.”  It was clever misdirection to put that 12% stat and interpretation first to give the illusion the stealthy “real people” premise had merit. It doesn’t. 

    #6’s analysis misses the mark most wildly, imo. This number is so high and the tweet count so low because of simple technique that follows the Indian maxim:  “Don’t poop where you eat.”

    In twitter, eating is monitoring for keywords, hashtags, celeb names etc. aka: “listening.” No tweet sending here. Don’t want these listening accounts “deleted, cancelled or suspended” because it’s just more hassle to keep listening. The only rule to follow is API load. Easy to program.

    Pooping is the sending of unpaid promotional tweets… wait for it.. based on matching the needs of the ‘client’ and the tweets heard by the listeners. Voila. The poopers are vulnerable to “delete, cancel, suspend.” The cost of doing “business.”

    Why doesn’t twitter stop all this spam stuff? Simple. It’s so much better to give the under-educated (tech and non-tech) media big numbers to report, no matter their true meaning, and let the misinterpretations con the public and the next target VC’s for the next round of money. Sure some poopers are shut down. So more poopers are made and the numbers  keep going up. See how that works? 


  2. gregorylent says:

    several million, if not tens of millions, twitter account have to be spambots or pornbots. 

  3. Bob C says:

    Hi Heidi
    Good dig.  Lots of great information. 
    A  gut check on depth of  Twitter penetration is to ask friends, family and colleagues.  In that regard the “10%  active”  across 500 mln users might not be far off.

  4. Twopcharts says:

    Hi Heidi, thanks for using our research in your post. I agree with you entirely when you state that behavior from new registrations cannot be extrapolated to the entire population, and I also believe that registrations from the last year will have less conversion to active users that the first 200 million registrations had. We will try to make a full estimate for active usage in the near future. 

  5. Venusbarak says:

    Smashing post Heidi, quite informative…

  6. Jim Campbell says:

    We believe that once Twitter themselves figure out who and what they want to be, it will be easier for people to see a value and participation will increase. Twitter is still a good thing, only going to get better. Thanks for writing this. It brought us to some thoughts we hadn’t visited for a while. Great job!!