There’s a particular statistic in the news this week because it’s been both widely shared and widely criticized. It comes from this Fast Company article on Apple. Here’s the key quote (emphasis mine):
Apple does an extraordinary job of extracting revenue from the worlds in which it already plays a role, and its future revenues will depend on this even more. Horace Dediu, an influential analyst now working with the Clayton Christensen Institute for Disruptive Innovation in Boston, estimates that Apple customers deliver an astronomical $40 per month apiece to the company, versus the pennies per month that Facebook and Google collect, and the few dollars a month that Amazon receives.
On Monday morning, I got pulled into a Twitter thread about this quote and the reasoning behind it but, because of Twitter’s character limit, I found it almost impossible to continue the thread in an intelligent way with six others (and their various usernames) taking up almost 90 of the 140 characters available. But it’s worth picking that quote apart a bit and looking at the actual numbers here.
Why revenue per user matters
Revenue per user (often abbreviated as ARPU for average revenue per user) is a much used metric in various services businesses because they generally know who their users are and can easily quantify them and know how much revenue each user generates. It is one of the two key drivers of revenue, along with the number of users itself. It’s generally much less useful outside of services businesses because others don’t have the same tight, predictable relationships between individual buyers and a regular monthly spend or revenue. If there’s growth in ARPU, that acts as a multiplier on user growth when it comes to revenue which is, of course, what all services companies aspire to.
Apple as a services company
Apple is not primarily a services company but a company that largely makes money through hardware sales. It describes itself as a company that provides hardware, software, and services in a tightly integrated fashion. Non-hardware items make up a small minority of sales at Apple, with just 10% of total revenues over the past twelve months. However, the company has been keen to talk up its Services segment and the revenue that flows from it, in part, because services are thought of as producing more predictable revenue streams than hardware sales, which have recently been less predictable as drivers of growth for Apple than in the past.
ARPU comparisons for Apple and online services companies
In that context, it’s inevitable people should start wanting to compare Apple’s Services segment with online services businesses to see how they measure up. Apple doesn’t report an ARPU number per se but for Apple, Twitter, and Google, we can come up with reasonable estimates while Facebook provides an official ARPU number. For Apple, we have two Services numbers we can use – one is the segment revenues reported under generally-accepted accounting principles (GAAP) and the other is its more recent gross revenue number for installed-base purchases (essentially that GAAP number plus the cut of App Store purchases that goes to developers, minus AppleCare+ warranty revenues). Note I’m using an estimate of Apple’s total user base which comes to around 800 million in Q2, on the basis that Apple’s total devices in use number of a little over a billion includes secondary devices for some subset of users, such as iPads and iPhones owned by the same users (I’ve assumed 20% of the total device base).
The chart below shows how some of these numbers stack up:
A few things stand out almost immediately:
- Apple’s gross services per user comes out on top in this peer group, with a little over $50 per year per user, while its net revenue (which is what it actually reports to the SEC) comes in quite a bit lower, and below the number for Google
- Apple’s revenue per user number is dropping but that shouldn’t be surprising – the new users that come on board are going to be spending less than existing users, because they’re disproportionately in countries with lower incomes and have fewer paid Apple services available (something I previously looked at in this post a couple of years ago)
- Twitter and Facebook are far lower on this metric, with Facebook in the mid-teens and Twitter still in the single digits per year, though both they and Google are seeing rising numbers
Given services aren’t Apple’s main business, it’s impressive it’s able to drive such significant spend simply off the back of its installed base of devices, especially as most of its Services revenue comes from services less than ten years old (notably the App Store, which debuted in 2008). If you were to take a step back and look at Apple’s total revenue per active user or device, you’d obviously get far larger numbers and those would obviously be much higher – you’re talking about somewhere between $250 and $400 annually over recent months, depending on whether you use the active device base or user base. But then you’re also talking about a less comparable business to those online companies.
To return to the quote we started with, if you’re being pedantic, you could argue Facebook did indeed generate “pennies” per month per user through 2015 (in that its annual ARPU was $11.96). So just under a dollar per month per user. But since the beginning of this year, Facebook has been generating over a dollar a month per user globally and Google has been well above that number for years now. On that basis, then, the Fast Company article is misleading.
US versus global revenues per user
Where things get really interesting is when you look at revenues on a geographic basis because the online companies generate dramatically different amounts in various regions and especially generate far higher revenues in the US than in the rest of the world combined. The chart below shows a US-focused split for the online services companies, with a couple of Apple’s global numbers for context:
As you can see, you get a very different picture when you look at things this way – Google’s US revenue per user is likely over $150 per year, while Facebook’s reported ARPU for US & Canada was over $50 for the past twelve months — roughly the same as Apple’s gross revenue per active user globally. Even Twitter’s US ARPU is almost up to the level of Apple’s global net services revenue per device. These numbers put the Fast Company quote in an even starker context – in North America, the online services companies generate much more than pennies a month and Google generates far more in revenue than Apple does in Services revenue per user, while Facebook is roughly on par with Apple. On the other hand is the trajectory, where Apple’s number has been falling as Facebook and Google’s numbers (and to a lesser extent Twitter’s) rise rapidly. That doesn’t actually matter that much from an Apple perspective because its installed base of devices and users continues to grow rapidly, driving overall Services growth.
No one way to measure a company
I’ll close with this observation: all these metrics are interesting to look at and can tell us something interesting about these companies, their business models, their growth trajectories, and so on. But these are fundamentally different companies, each of which has unique characteristics that make single metric comparisons overly simplistic (my post last week compared some of these companies and others on another set of metrics, for example). Yes, it’s interesting as Apple sells itself more and more as a services company to put those claims into context, but the reality is Apple as a company is still ten times the size of its Services segment. It’s far less dependent on ARPU growth for its overall revenue growth than are pure-play services companies. Regardless, Apple doesn’t need observers to make false comparisons of its revenue per month per customer with other companies – I’m pretty sure Apple can handle being compared accurately with its competitors and peers.