Earlier this week, Apple wowed us all with bold claims of a 50% adoption rate for its brand new iOS 9 software update.
The revelation meant that iOS 9 – the iPhone 6S launch software – represented the fastest iOS adoption in history, but do the figures stand up to scrutiny?
According to Apple, its data was recorded based on App Store visits on September 19 – three days after launch.
However, information collected by a number of data research companies appears to refute the 50% adoption claim.
For instance, both Crittercism and Fiksu measured 23% adoption on September 19.
MixPanel, another established data analytics company, noted the figure at a slightly higher 29% – still far below Apple's purported rate.
Related: iOS 9 Features: What's new?
“When you look at the three of our numbers...Fiksu and us are identical, or almost identical. MixPanel is a little higher, but all within what would be considered a normal statistical margin of error,” said David Appelbaum, VP of Marketing at Crittercism, speaking to TrustedReviews.
But Appelbaum isn't the only one to point out the strange data discrepancy. Suhail Doshi, MixPanel's CEO, also noted the contradiction, tweeting:
How do they collect their data?
According to Appelbaum, Crittercism accrues its data by embedding analytics tools into various popular apps on iOS.
This list of apps includes LinkedIn, eBay, CNN, Fox, Disney, Etsy, Groupon, American Express, and AT&T, to name just a few.
Crittercism then builds a picture of adoption rates based on app loads and app usage.
The method is a popular approach to researching adoption, and Appelbaum tells us it’s been very accurate in the past.
“Particularly as we go back historically, to say iOS 8 adoption, various flavours of Android etcetera, we’ve found it to be remarkably accurate,” the exec tells us.
For example, iOS 8 released on September 17, 2014, and crept over the halfway mark for adoption on October 20, registering 52% according to the App Store.
Crittercism's data put iOS 8 adoption at 48% on October 17, 2014, which appears to fit with Apple's own data from the period.
So when Apple announced its startling adoption rate so early on, the Crittercism team was very surprised.
The shock stemmed from the fact that as of September 19, Apple hadn’t even pushed global notifications about the software update to its users.
“They have not started their active push down yet. You know, the push notifications that say you have an upgrade available,” explains Appelbaum. “So a significant part of the population isn’t even necessarily aware that there’s an upgrade yet, since they haven’t actually hit the push button.”
Appelbaum continues: “So the assumption that there’s 50% adoption when they haven’t yet gone to active push really makes no sense at all. We’re shocked, actually. It just doesn’t fit, right?”
Why is there such a big difference?
Why is Apple peddling a 50% adoption rate when research suggests otherwise?
Crittercism’s marketing lead believes that it's all about how Apple collects its usage data.
“We think that it was probably based on App Store traffic. So our assessment was that they probably based their upgrade stats on folks who were coming to the App Store, as most likely candidates for upgrade, and extrapolated based on that,” Appelbaum reveals to TrustedReviews.
The suggestion here is that regular App Store users are more likely to update their software compared to those who don’t use the store regularly.
As such, adoption by regular App Store users would be significantly higher than the actual adoption.
By choosing this method of data collection, Apple may have inflated its adoption rate to more than double that of the actual figure.
Fiksu, another data analyst firm, agreed this may be the reason for the inflated figure.
"We think there's a pretty simple reason. Our data is gathered from app usage – apps that have our SDK are installed on hundreds of millions of iOS devices, and every time a user launches one, we get that event as a data point, and it includes basic information about the device including the OS version," explains Fiksu's Jeremy Sacco, speaking to TrustedReviews.
Fiksu's iOS 9 adoption data
Sacco continues: "Apple's data comes from visitors to the App Store. The set of users who are actively adding apps skews more active – they're more likely to be updating their OS. Our data represents just basic usage of phones (and tablets), so gets a lot more of the casual users who don't frequently add new apps."
There are some other possible explanations as to why Apple’s adoption rate figures might be so different.
For instance, Apple could be reporting the percentage of eligible users that have upgraded to iOS 9. Remember: the software is unavailable to phones launched prior to the iPhone 4S.
However, Crittercism reckons this isn’t likely. That’s because the company filtered out old devices that don’t support the software, and saw the adoption rate rise from 23% to just 25%.
Those figures were trialled in both the US and globally, and the numbers stayed the same.
Speaking at the time of the Apple announcement, Phil Schiller, Apple’s senior VP of Worldwide Marketing, said: “iOS 9 is also off to an amazing start, on pace to be downloaded by more users than any other software release in Apple’s history.”
Unfortunately, it’s now tough for us to know exactly how well Apple’s latest mobile OS is really doing.
We’ve contact Apple about these discrepancies, and will update the article with any response.
Do you think the 50% adoption rate is too good to be true? Tell us your thoughts in the comments.
Check out our iOS 9 features video below: