How a small change by Apple cost Twitter millions of users

Getting people excited about it again
Getting people excited about it again
Image: Reuters/Kacper Pempel
We may earn a commission from links on this page.

Twitter says it lost 4 million monthly active users last quarter after Apple made changes to iPhones and iPads. What, exactly, happened?

The preponderance of those users—about 3 million of them—were lost when Apple’s mobile web browser Safari stopped automatically requesting information from Twitter’s servers, CEO Dick Costolo said on a conference call with analysts. He didn’t explain further, but was undoubtedly referring to the “shared links” feature, a backwater of Safari that displays links shared by people you follow on Twitter.

It looks like this:

Image for article titled How a small change by Apple cost Twitter millions of users

Apple first integrated Twitter with its mobile operating system, known as iOS, with the fifth version in 2011, a deal that helped boost user growth by exposing the service to more people. Two years later, with iOS 7, Apple made the connection deeper by adding shared links.

To make the feature speedy, Apple would periodically check Twitter for the latest links shared by people you follow. It didn’t matter if you were a Twitter addict or just signed up for an account one day, connected it to your iPhone, and promptly forgot about it. Because Apple kept checking your account for data, you would count as a monthly active user in Twitter’s statistics.

But when Apple released iOS 8 in September, it also changed how Safari’s shared links work. It no longer automatically asks Twitter for information, instead updating only when you actually open up the feature. Do it now, and if you are using iOS 8, you’ll see it takes a few seconds to load. (Oh, and if you do that, congratulations, you are now one of Twitter’s 288 million monthly active users, or MAUs.)

The issue of whether Twitter should count those kinds of users has come up before. In its latest earnings report, the company said “8.5% of users”—that’s 24.5 million people—“used third party applications that may have automatically contacted our servers for regular updates without any discernible additional user-initiated action.”

So those users probably shouldn’t count as MAUs, at least to investors trying to gauge how quickly the company is growing. Twitter, meanwhile, has been trying to convince Wall Street that MAUs are a poor measure of its size and health. It has a strong case, and investors seem to be coming around on the issue: Twitter’s stock was up more than 10% in after-hours trading after it reported anemic MAU growth but very strong revenue figures.

On the conference call, Costolo said that, in addition to the issue described above, another 1 million MAUs were lost last quarter due to “an unforeseen bug” in iOS 8 that Apple has since fixed, stemming those losses.

Update (Feb. 6): In a tweet the following day, Twitter walked back its claim that the second issue, a bug that cost it 1 million MAUs, was Apple’s fault. “There was no bug or issue with iOS 8,” the company wrote. “It is an issue on Twitter’s side as users upgraded.” In an interview with Business Insider, Costolo described it as an “encryption issue” but didn’t explain further.