Twitter’s long-term traffic still growing

Twitter users, most of whom signed up for the site and then apparently stopped visiting, are actually coming back, new data from comScore suggests. There’s been a lot of evidence that Twitter’s growth is tapering off by several measurements: monthly unique visitors, searches, and mentions in the media. But as VentureBeat guest columnist Jesse Farmer postulated in May after crunching a lot of Twitter data, the mass of people who aren’t early adopters could just be taking a while to warm up to using the service.

Today’s comScore numbers back up that theory. At first glance, Twitter’s rocket-launch growth curve seems to be leveling off. Worldwide traffic measured 37.3 million people in May, a 16 percent rise from April. That’s far less than the 68 percent growth from March to April. In May, its US traffic grew even less: Only 3.5 percent, to 17.6 million.

But as Erick Schonfeld shows in an excellent analysis of comScore’s numbers, other new data indicates a bigger story. Worldwide pageviews came in at 900 million for May, a 30 percent increase over April, with the average user now logging on almost daily. And this traffic was partly driven by people coming to read tweets on the site, not necessarily to tweet themselves. As he puts it:

A lot of [new users] come back, either because they keep getting links from friends or keep hearing about it on TV or whatever, and then they slowly start to see the usefulness—a funny Tweet from a friend, a link to breaking news, a way to keep an eye on the general zeitgeist. Twitter is the kind of thing that is easier to experience than it is to explain. But it is an acquired taste and often requires repeated exposure before people get hooked. Once they do get hooked, there is no going back.

That is exactly what happened for me and to many friends who signed up for the service in 2007 — we didn’t really start using it heavily for months, and now we do.

ComScore’s numbers today side with Farmer’s earlier thesis. In his article, he took issue with a report by market research firm Nielsen that said first-time users didn’t seem to be coming back and tweeting. He said it was too early to tell, and that it could quite likely be that Twitter users just weren’t active, and just hadn’t gotten used to the service. Here’s a big, juicy excerpt from his post:

Nielsens’s report equivocates about what, exactly they are measuring. Quoting David Martin, Vice President, Primary Research at Nielsen Online: “We still maintain that the majority of people who sign up for Twitter won’t be around in a month.”

This isn’t even what their own report says. Their report states that 40 percent of users fail to return to Twitter in the month immediately after they join. It says nothing about the second, third, or fourth months after they join.

For example, if I join Twitter in January but don’t tweet until March I would be considered a “Twitter quitter” in Nielsen’s report. This behavior is common on immature social networks where new users who join don’t yet have any friends or followers.

It also turns out that a large portion — over 20 percent — of the people who fail to return the next month never once updated their status. It seems unfair to penalize Twitter for users who sign up and never tweet. After all, one could just as easily pretend they never signed up and the retention rate would increase by 50 percent in April alone, from 40 percent to 60 percent.

Accounting for these two failings in Nielsen’s report dramatically changes the picture of Twitter’s “dismal” retention.

So, Nielsen’s numbers may be technically correct, but they miss the bigger picture and unfairly penalize Twitter. These numbers are actually conservative because we only count users who return and tweet in our retention rate, while Nielsen also counted users who returned but never tweeted.

We still don’t know Twitter’s long-term potential — maybe it will not become popular outside of the celebrities, journalists, marketers and other people who are crowding the service today. Or maybe things like its role in major events — such as the popularity of its real-time search during the Iran election crisis — will make it more and more mainstream.

None of this, I realize, answers the long-standing question of how the company will ever make money. Many of the things it could do to bring in revenue, other sites could do just as well. Advertising paired with real-time search? Google could do that using Twitter’s own data. Premium tracking and management features? A whole raft of social media startups are already at it.

The risk Twitter runs is that by letting other companies handle expensive, complicated or risky services for its members, it leaves itself with little  left to sell. There’s an opposing theory that by letting those companies offload the high-cost services, Twitter will win by staying lean and focusing on its core stream of tweets around which an entire ecosystem has grown (and at some point find a great new revenue stream).