Yesterday, Nielsen Online ran an entry on the Nielsen Wire blog titled Twitter Quitters Post Roadblock to Long-Term Growth. In it, David Martin, Nielsen Online’s Vice President of Primary Research, claimed that on a month-to-month basis Twitter fails to retain more than 60% of its users. Today, everyone from writers at Business Week, PC World, and The Christian Science Monitor to social media marketing bloggers trumpeted the story as proof that Twitter is nothing but hype. But upon closer inspection it appears that Nielsen Online’s post was either ignorant or irresponsible.
In the blog post, Martin provides no documented statistics to back up Nielsen’s claim, nor does he give any indication of Nielsen’s methodology. The term used, “audience retention,” suggests that Nielsen measured page views or site visits rather than actual use of the micro-blogging service. Martin then compounds this problem by comparing the Twitter retention rate “statistics” created by Nielsen to similar statistics for Facebook and MySpace when they were at an audience level similar to Twitter’s, and suggesting that Twitter has not succeeded at the same level as the two social networking sites.
But there’s an enormous problem both with this comparison and with Nielsen’s methodology: unlike MySpace and, until recently, Facebook (which announced an Open Stream API this week that allows users to access their feed information via third parties without going to the Facebook website), Twitter users don’t need to visit the Twitter website to engage with other users. In fact, according to the Twitter analytics service TwitStat, less than 1/3 of Twitter users access their accounts via Twitter’s web interface.
One of the unique things about Twitter—and a driving force behind its explosive popularity—has been its willingness to allow users to access their accounts via third-party applications such as Tweetdeck and TwitterFeed. According to TwitStat, Tweetdeck alone accounts for nearly 20% of all Twitter usage. In other words, judging by Nielsen’s post, a given Twitter user (such as Molding Web) that accesses his or her Twitter account via Twitter.com and then switches to using Tweetdeck would be included as a user that Twitter failed to retain. Such an assumption is not merely misleading; it displays a total ignorance of the fundamental appeal that Twitter has to smartphone users and other hyper-connected professionals.
