Are Twitter’s Ghost Users a Problem or an Asset?

Ghost descending the staircase at Raynham Hall, Norfolk, England. (Photo by Time Life Pictures/Pictures Inc./The LIFE Picture Collection/Getty Images)
We have tons of users like this guy, we swear. Photo: Time Life Pictures/Getty Images

A persistent knock against Twitter, by the people who care about these things, is that it’s not as big as the competition: only about 100 million daily users, compared to Facebook’s 800-odd million. The counterpoint to this observation has always taken the form of an existential question: What is a user, really? If someone reads a Huffington Post story that contains an embedded tweet, is that person using Twitter? If a CNN viewer sees Anderson Cooper reading tweets about Gaza aloud on his show, isn’t that person part of the Twitter audience, even if she’s not personally logging in?

This disconnect has created an odd issue for Twitter: It has no idea how many of these “ghost users” it has. Ghost users don’t log in to Twitter, so you can’t count them in the traditional way. And even if you could measure the number of ghost users, it’s hard to make money from them, since embedded tweets and tweets shown on TV don’t contain the same ads that run on Twitter’s main site. 

Yesterday, after reporting solid quarterly earnings that sent its stock up 30 percent, Twitter again argued that while the size of its passive, non-logged-in audience can’t be measured, it’s there, and it’s big. Big enough, it contends, that even though the company may not reach a Facebook-size audience directly, Twitter has seeped into the cultural lifeblood in a much bigger way than traditional metrics can capture.

Twitter CEO Dick Costolo was especially emphatic on this point:

Twitter is everywhere,” Mr. Costolo said. He claimed this audience was “two to three times” larger than the monthly active user numbers. And he noted that the company was experimenting with ways to “improve the content” for the visitors who are not logged into the platform. Still, Mr. Costolo insisted that ad products are not being sold or considered to reach these visitors. “We are focused one-hundred percent on the user experience today,” he said. “We’re not monetizing those audiences.”

Of course, Costolo has no idea how big “those audiences” are. That’s the point. It’s unknowable, like trying to measure the number of people who will hear a Katy Perry song in their lifetime. You can measure the traffic to, but that doesn’t include the number of people who see tweets on TV, in the newspaper, or on sites like Reddit, where users routinely screen-capture popular tweets and disseminate them to millions of people, none of whom need to be logged in to Twitter to see them.

That epistemic impossibility must kill Twitter’s executives — what they wouldn’t give for a true measure of their importance! — and it’s also hobbled Twitter’s ability to sell ads to companies, who like metrics and micro-targeting. After all, if you’re Coca-Cola, do you want to show your ads to the Twitter audience you can’t measure, or the Facebook audience you can? This is why Twitter may be developing new metrics to try to capture the size of the ghost audience — to show advertisers, somehow, that their dollars are being well spent.

But the ghost-user problem is less important than it once seemed, in part because Twitter has done a good job of maximizing the variables it can measure. (As Will Oremus puts it, “Twitter did so well on its existing metrics last quarter that there was really no need for new ones.”) Revenue is up 124 percent in the last year, and the company is nearly breaking even, in adjusted terms, an improvement over quarters past. By selling ads more effectively to its existing users, Twitter has managed to make itself significantly more valuable, even if the overall usage of the service isn’t growing much. And advertisers seem to be content, for now, knowing that Twitter is entering the conversational lifeblood, even if they don’t know how widely. In the span of a single quarter, the ghost audience has gone from being Twitter’s greatest liability to being an extra for investors; buy Twitter for the users it knows about, the argument goes, and you get the ones it doesn’t know about for free.

Of course, relying on unmeasurable growth is a dangerous game in the long run. The goal, if you’re Twitter, is to get so big that it doesn’t matter how big your overall audience is — just the number of users logging in is sufficient to command a huge market value. But until that happens, Twitter is smart to focus on its top- and bottom-line numbers, the variables within its control, while relentlessly emphasizing that the ghost users are there. Even if Wall Street can’t see them.

Twitter’s Ghost Users: Problem or Asset?