Android vs. iOS: Revenue Per User Won't Catch Up and That's Okay

The metric of choice for comparing iOS to Android is average revenue per user, or ARPU (spoken phonetically exactly as you'd guess). This is fine, it is certainly technically accurate, but it isn't exactly relevant for most of the discussions where it is used as ammo. It's like comparing the average revenue per user of all car buyers vs. only Ferrari* buyers – it tells you what you already know.

Notice how I didn't say Hyundai vs. Ferrari, that would be a hyperbolic falsity, I specifically said all cars because that's much more aligned with the reality of Android. You need to understand the Android spectrum to fully appreciate the volume, human impact, and glory of what they've pulled off.

This graphic shows the ARPU comparison between the platforms, it is interesting data, but remember to keep perspective on the topic (source article).

Android vs. iOS ARPU vs Deepak Abbot on Medium

Now some considerations and perspective:

  • Note the ARPU difference and relationship in app sales vs. ad revenue. 
  • Android has been installed on a lot more phones than iOS. I am basing this on data that is getting old, but I don't think enough has changed to change the blunt fact that Android dominates on pure install base – Android had (Q4 2013) 78% market share to iOS's 18% worldwide. The data above shows a 74/26 split purely between these two platforms. Before you get angry at the mention of install base metrics...
  • Install base doesn't mean much of anything for any discussion aside from install base discussions.
  • There are many millions of Android device owners that don't own a computer or have another source of Internet in the home, the phone is their first and only computer and their connection to the web.
  • There are Android users who don't have running water, plumbing, or electricity at home. Communities have communal charging stations on the side of the road. Benedict Evans does a great job of tracking and tweeting (or retweeting) data and anecdotes about this. Can you imagine walking a day or more just to charge your phone?
  • Android devices are connecting people from the third world to services and communities that allow them to catapult their communications capabilities forward by decades.

There isn't one smartphone market. You can make it look like a single market on paper, but the reality is there are several markets. I would love to see data on premium device ARPU broken out by region, this would be a more valid comparison to make. There are hundreds of millions of Android devices that are in the same premium segment as iPhones, and there are old hand-me-down iPhones that are in a similar segment to lower end Android (though I don't believe even the bottom of the iPhone spectrum reaches the price points/capability of the low end Android phones). These devices have changed the world overnight unlike anything most of us have ever seen.

Next time you pull out your device of choice and think that you can't live without it, remember that there are millions of people for whom their smartphone is fundamentally changing the entire trajectory of their family's livelihood. Apple and Google have enabled extraordinary things with their platforms and there is no question about our love for and dependence upon our phones. However to compare the two platforms outright as one market is a fool's errand without maintaining perspective on the extraordinary differences between what the platforms (or even subsets of each platform) enable.

*I picked Ferrari because I originally picked Audi and then started questioning whether the revenue from super cars would actually be enough to offset the lower end cars and render my analogy dead in the water. Rather than picking Audi and doing the math, I went with a car I knew would represent a huge ARPU that the car industry as a whole couldn't touch.

Two Factor Authentication Matters - iMore's Guide to 2FA

I am a firm believer in two factor authentication; I can't imagine a better way to keep yourself secure than enabling 2FA on any service that supports it. Contrary to popular belief, there is minimal inconvenience. For any site or service that you care about, the benefits outweigh the cost by several orders of magnitude.

It's worth taking a look at who all supports it and noticing the pattern that that so many "old dogs" (companies that are huge and have been around for a long time) don't support it, but newer companies do. This shouldn't come as a surprise, but might help you decide who gets your business.

The Safe Mac

If you use iOS or OS X and missed the Mac Power Users podcast this week with Joe Caiati, I highly recommend checking it out. One highlight for me (as the default tech support for a growing circle of people) was the mention of The Safe Mac. The website is "retro" (to put it nicely), but the adware removal tool is top notch – and free (donations recommended). 

It is worth checking out as an option for your loved ones who find themselves installing Flash from some bogus website (something that's shockingly easy to do).

Twitter's New User Problem

Twitter's growth numbers have been a disappointment among investors and a large group of the company's enthusiasts. On one hand, having hundreds of millions of users is an incredible accomplishment. On the other hand, they are having tremendous difficulty bringing new users onto the service and keeping them there, especially for a company whose brand recognition is off the charts.

There are a handful of fundamental issues with both the perception of the service and the new user experience that will prevent this reality from changing. Make no mistake, I am bullish on Twitter and wouldn't think twice about declaring it my favorite thing on the internet, but these roadblocks must be overcome to see any meaningful change. Twitter has the potential to be the central hub for how people experience the digital world around them. It is social, it is news, it is entertainment, it is educational, it is anything and everything you want it to be. So why aren't more people signing up and sticking with the service if it is so versatile and powerful?

The Public (Mis)Perception of Twitter

Twitter has completely lost control over the perception non-users have of the service. People see hash tags on TV shows, periodic mentions of the company on the evening news, and sometimes a scandal or two. To someone who doesn't actively use Twitter, they can't even define what it is for or why they should use it. I frequently ask non-users why they aren't on Twitter and the response is always (I literally mean 100%) some version of, "What would I use it for?"

It is difficult to portray your company clearly when it is so many different things to so many different people. A cohesive message about the service they provide will either only apply to a subset of users, or it will be so miserably vague you will end up with something like Twitter's official Strategy Statement. Even asking the most enthusiastic users what Twitter is doesn't consistently yield a reply that inspires people to sign up. You can describe it as I did earlier, "It can be anything you want it to be," and scare them away because it sounds overwhelming. You can describe it with specific examples that are in line with the inquirer's interests or career and you might have more luck, but then you have to tailor an answer to every person and you will give them a skewed perception of what Twitter actually is. You can say, "Let me show you!" and quickly learn that your Twitter feed is probably only interesting to you while you overwhelm them rapidly navigating the lingo and stream. 

I don't pretend to have the solution to this problem, but I can say with confidence that there is tremendous power in giving potential users the perception of your company that you want them to have. Consider the Apple Watch event where the most common criticism is their lack of their famous "why you need this product" speech that has so famously been an integral piece of all new product releases from modern Apple. 

When users don't understand the need for what you are offering, you're going to have a very tough time getting them to spend time, money, or effort on your product. Twitter is overwhelming, but it doesn't necessarily have to be. Perhaps by fixing the daunting onboarding process Twitter can gradually change the public perception by converting the uninformed into enthusiastic users.

Twitter's New User Experience

Along with the breadth of possibility comes the abundance of complexity and ample confusion. If you can get people to sign up, which is a big if, they now have to navigate turbulent water as they build their network. Start with a few friends, maybe a favorite TV show or sports team, sprinkle in a couple of hobbies, and they've at least got something to look at.

Let's be honest though, the official Google account isn't the interesting source on Twitter for all things Google. Likewise official sports team accounts are usually boring and generic. Products, services, or TV show accounts just spam you with promotions and prompts to retweet content and spread the word. This isn't what makes Twitter great, not by a long shot.

There needs to be a smarter, simpler, and much faster way to show a new user valuable content that they care about. It seems safe to assume that most people who don't use Twitter today would sign up with skepticism, if they signed up at all. This means that every moment matters, every fraction of a second works against you if you aren't providing value. This is the new user experience, and right now Twitter is failing in a big way. 

There is an opportunity to automate this initial experience by leveraging one of the largest data sets in the world to algorithmically determine what a user might be interested in. With some kind of seed to the profiling algorithm Twitter can present the user with high-level topics; then as selections are made, narrow the focus slightly. From here, you give the user a sense of the tweet volume and allow for adjustment via a slider or perhaps even automatically. Based on the frequency the user launches the app, what they do in the app, and how they have it open, you tailor the content in the stream to appeal to them. Very quickly, assuming subtle UI hints (not tutorials) show the user how, the user will hide the auto-followed accounts if they don't like them and start following accounts they do like – both of which improve the algorithm. Depending how good it is, you could even omit the tweet volume slider to simplify the process, just present great tweets on a valuable topic. If you're providing tangible value to the user, they will want to curate a list of people to follow, they will want to spend time on the service getting more value, and they will want to spread the word.

The fundamental point is that in less than 30 to 60 seconds, the user needs to be immersed in Twitter, and not just the app, but a version of Twitter that is valuable to that individual user. The friction experienced between launching the app and gaining value from it will overwhelmingly determine whether this person will remain a user. With every tiny hiccup, extraneous question, or mundane tap the user has to indulge for the sake of Twitter rather than themselves, the less likely they are to stick.

I mentioned it briefly above, but there will need to be some kind of seed for this algorithm. It doesn't seem likely that you can safely rely on people linking their Facebook profile, and even if they do link it, most people use Twitter so fundamentally differently that you don't want to scare them into thinking this is just a weird version of Facebook. You could ask for access to their contacts and seed it with their friends and who their friends follow, but that isn't a golden ticket either, not to mention the growing disdain for allowing access to your contacts within apps. This strikes me as the hardest problem to solve without having the user fill out some type of survey of their interests, which not only takes time and effort for no immediate apparent value, it's just a generally terrible first launch experience.

Maybe there is some solution in the huge pile of data Twitter has for this, or maybe not, but even a slightly uglier sign up process that generates a meaningful stream of content is better than the new user experience today. There are services like WeFollow and Twtrland that can help you to find people to follow based on search terms, but the average user won't invest this much in building a follow list until they've uncovered the value they get from their effort, if they ever do.

[UPDATE: Twitter has improved this process over the last year or so, but it still feels generic. I am pleased to see a very brief survey of the user's interests yield some suggestions that extended beyond the boring major brand accounts. Unfortunately, at no point during the signup process do they attempt to give the user any kind of idea what Twitter can be to them; nor is there any automated way to refine the list of who a user follows based on their activity. This is definitely a step in the right direction, but not enough to change the minds of people who have previously written off the idea of using the service.]

Now What?

Experienced users have a list of complaints and requests about the service too, many of which I would agree with. However, the public perception of Twitter and the onboarding problem strike me as two major flaws that most directly impact the growth outlook for Twitter. 

Clearly this isn't a simple solution, or Twitter would have done it already. However, there are very basic and obvious roadblocks for the new user experience that prevent Twitter from growing like investors want and enthusiasts imagine. Twitter is one of the greatest services and inventions in recent history, but they're far from realizing their full potential. The overwhelmingly emphatic supporters of Twitter, like myself, are large in number and I believe highly justified in their beliefs, but we seem to be running up against the limit of how widespread the service can get without addressing these basic concerns.

Messaging: Speed or Consistency, Not Both

Messaging is king, especially text-based messaging. Voice is fine, video has its place too, but there is nothing today that can replace the rapid and highly private aspects of text-based messaging – no headphones required. Given this, it is no surprise that iMessage has been a hit, but other cross-platform apps have been even bigger hits (judging by user count). WhatsApp, Telegram, Hangouts, and others are all trying to be the go-to app for messaging.

This isn't a roundup or definitive review. I want to address the (admittedly first world) problem of getting speed or consistency, but never both. What do I mean by that? Let's take iMessage and compare it to Hangouts for the demonstration, but "Hangouts" could just as easily be replaced with WhatsApp or others.

iMessage – Brilliant but Inconsistent

When you open iMessage you immediately get your conversation content because it is stored right there on the device. There is no permanent copy on the cloud (details here). Messages are delivered securely to a device and that's the end of it. The result is a lighting fast feel and a negligible launch time, it doesn't even have to ping a server until you send/receive a message or compose a new one (it checks if they're an iMessage user).

Unfortunately, with iMessage you also get inconsistency if you have more than one device with iMessage activated for the account. When my wife sends me a message and she sees it gets "Delivered" she has no idea which of my devices actually got it. You would hope they all did, but that isn't always the case. Technically each iMessage results in the server having N secure and isolated copies where N is the number of devices on your account. Each copy is treated independently and is only removed after it has been confirmed as delivered or it reaches the 7 day expiration. Although this is how it supposedly works, in practice there must be some logic that only confirms a message was delivered to a device and after a certain (very short) period of time it treats all N copies as delivered. Far too many times have I come back to my iPad after a day of missing iMessages on my phone only to find them on my iPad for my assumption to be patently false.

I will stop here and say iMessage usually handles this correctly (far better than it used to), but not always, and not everyone's experience is as good as mine. Messages showing up out of order or some messages never showing up at all are all common problems if you've got more than one device connected to an iMessage account.

Hangouts – Slow and Steady

Hangouts, on the other hand, never shows messages out of order (or effectively never) and they always show up. Sure you might not get a push notification, but if you open the app they will be there. The main reason behind this is that the master copy of a conversation is on the server, not the device. So if you never get a push notification for a message (which is annoying) the server likely thought you read that message on another device already. This model of having the server as the master means that every time you launch the app it has to ping the server and download updates. Even if you got a push alert with a message, when you launch the app it still pings the server and re-downloads that content that it just showed you in the alert. In iOS you can have a push notification trigger a background download of the message content, but Hangouts doesn't take advantage of this, much to my discontent.

It seems small, perhaps even insignificant, but given with text-based messaging I want to be in and out as fast as my fingers can tap. I will often go in and out of conversations many times in a single minute. Hangouts, despite only requiring a few seconds of my time, feels unusable as an iMessage replacement because of the lag. WhatsApp has similar launch lag and every new message has to be downloaded when the app launches instead of being ready for me when I open the app, unlike iMessage.

Note: I know this launch and usage experience is different on Android, it is far less of an issue there. I still believe iMessage to be a superior experience within the limits it imposes, but it is worth pointing out that these shortcomings are based on the iOS app.

Solving the Problem

There is a level of complexity to this problem, as simple as it may sound. You don't want to push alerts to 3 devices for every single message if a user is carrying the conversation on one of them. At the same time you don't want to miss message delivery entirely. None of this even matters if the user experience isn't good enough for the messaging app in question to gain the necessary popularity.

How do you solve it? Keep a master copy on a server? Have a "primary" device option? Have behind-the-scenes read receipts? 

I think a combination of fail-safes built on top of the current iMessage strategy is likely he way to go. Avoiding the server-as-the-master solution is the only way to get a truly fast experience. If there were logic to actually check for per-device delivery as well as read receipts  for the server (without having to turn on actual read receipts), most of the issues seen with iMessage today would likely be avoided. It would also make sense to have a primary device either automatically determined based on usage or manually configurable. This could mean that with 100% certainty a single device would always get every message no matter what. There are issues with that device not being connected, but there are solutions beyond the scope of this piece.


Text-based messaging is private, convenient, and fast. The private nature makes it intimate and on a personal level makes us trust it and love it in many cases. It is fast, it is reliable (mostly), and it is simple. Anyone can do it regardless of their device or platform; and while we're seeing the proliferation of apps and methods, it has never been more convenient to use text-based messaging to communicate. I believe that the speed and reliability of the experience are key which is why I vastly prefer iMessage to alternatives, the above average level of security helps too. With that being said, I think iMessage has a long way to go for me not to caveat my otherwise emphatic endorsement of it.