A Wall Street Journal post from two years ago shows a list of apps that gather your data and what you do with them. Angry Birds not only collects your contacts — it transfers them to third parties.
There were apparently people that were shocked — shocked — to find that Path, the slick, occasionally pretentious don’t-call-us-a-social-network social network was uploading your iOS address book to Path servers.
Among Path’s errors was not notifying users of this, but as Mugunth Kumar has pointed out, uploading your address book to third-party servers is hardly new. Hipster, FourSquare and the über-beloved Instagram all do the same thing, with various levels of skeeziness.
Given the choice, none of these apps would ask for permission. After all, that just increases friction in the sign-up process and these startups have read every possible thread on Hacker News about making user acquisitions as smooth as possible.
But even among the apps that do ask for permission, the question they’re asking doesn’t match what they’re doing. WhatsApp says “‘WhatsApp’ Would Like To Access Your Address Book”. Viber has the slightly more forthcoming “The Viber service needs to access your address book, and sync your contacts list.”
The harm in these notifications is that they sound like they just want to take a quick peek at your contacts and see if any of your friends are using the same app. Of course, that’s not what they’re doing, or even what they want to do. Users probably wouldn’t want this either, as it would make things a lot more difficult to find out when a friend of yours joined Instagram.
But if they were honest these notifications would say, “We want to take your address book and upload it to our servers.”
A message like that is probably going to make people a little more hesitant before pushing “Okay” with their thumb.
Then we’d hit “Yes”. We always do.
Horace Dedieu has the breakdown, and the third chart is just astounding. What’s even more remarkable is that Apple is selling less than 10% of all phones. As I’ve said before, this is good for Apple but bad for consumers. It’s not just that we need more competition — we need more profitable competition. Samsung is the next best hope, but they’re too busy cribbing off Apple to hope that they’re going to build anything sustainable.
So the question is this: is Facebook going to be as big as Google?
Is Facebook going to be bigger than Google?
I’ve got my doubts, after looking at their S-1. Facebook’s got to book numbers like the $38 billion in revenue Google posted in 2011. There’s no denying that Facebook is in good shape, but there are still two key issues facing the company: ad quality and mobile.
Ad Quality
Ad quality is a component of how much money a company can make off an ad, and it reflects how valuable the ad is both to the advertiser and to the viewer. The higher the price of the ad, the higher its quality. It’s the reason Google’s been able to make so much money: searches tend to be tied towards purchase intent. Facebook’s betting that targeting is as important as intent.
Let’s get started by looking at how much money Facebook could be making. For the moment, let’s limit this to the U.S. market. People here spend more than four times the amount of time they spend on Facebook than they do on Google. But Facebook only books $2.07 billion in domestic revenue — less than eight times the $17.56 billion Google turned over in 2011. So even though users spend far less time on Google than they do on Facebook, advertisers are still willing to spend more with Google.
That points to poor ad quality and is exactly why Facebook is rolling out things like frictionless sharing and “Featured” ads. They know that they have to be making more money, but they just haven’t figured out how yet.
If they want to be making as much money as Google does in the U.S., Facebook has to be making over 8 times what they’re currently making. If they want to be making as much money as Google is making per user minute they have to making 36 times that amount of revenue — $72 billion a year from the U.S. alone.
$72 billion in domestic sales sounds ridiculous — those are Apple numbers — but is it possible for Facebook to increase their revenues by eight times what it is now? Google managed to increase their earnings by roughly 10 times after going public, so it sounds doable on paper.
Facebook has already started to do this by increasing the value of their ads shown to consumers. The way to calculate that is by looking at the number of Monthly Active Users (MAU) and compare it to their revenue numbers disregarding growth. In 2010 Facebook had 154 million MAUs in the U.S. and Canada, and booked $1.2 billion in domestic revenue.1 That works out to roughly $7.9 million in revenue per million MAUs. In 2011, Facebook had 179 million MAUs in the same region. If their ad quality hadn’t improved, Facebook’s revenues would be about $200 million higher than it was in 2010, or $1.4 billion. Instead, it was $2.067 billion, a significant increase. Facebook managed to increase their revenues per million MAUs number to $11.54 million. Which means that they were able to improve their ad quality by 45%.
The issue is that Facebook’s ad quality growth is slowing. In 2010 they were able to increase their ad quality by 71%. The 2011 45% number is still great — if they’re able to hold that constant they’d reach Google’s U.S. revenue numbers in five years. But they’re not going to be able to sustain that — it’s already slowing down — which raises the question of how fast Facebook can ramp up ad quality, user growth and engagement to withstand the competitive pressures they face from other potential challengers and time sinks. How quickly do they need to do this? Well, we’re already seeing some clues that Facebook’s user engagement is decreasing, like the decline in online videos viewed on Facebook. Which brings us to the second issue.
Mobile
Facebook hasn’t turned on their ad solution for mobile yet though it’s clear that they’re going to, and probably quite soon. Out of Facebook’s 845 million MAUs, 425 million were using Facebook mobile products. The proportion of mobile users is clearly going to increase over time as more and more people shift away from traditional desktop browsing. The issue is how much money Facebook can make off of mobile users.
As Google is finding out, it’s not easy to make desktop dollars in the mobile market. Mobile search is less profitable than desktop search, as Henry Blodget noted. Google and Facebook aren’t strictly analagous here, but it’s clear that there is a core difference between the advertising you can make off traditional web products vs. mobile products.
More than that, there is also a fundamental difference in the type and amount of advertising users are willing to put up with in mobile. Facebook’s core product — their desktop web experience — can’t be shrunk down and slapped into a mobile product. Everything about the way they approach their product has to be re-thought out, especially on how they’re going to serve the amount of ads needed to justify their valuation. Is mobile ad volume going to be anything like desktop? Doubtful.
And then there’s the Platform problem. Facebook generates 15% of it’s revenue from Platform revenues — with Zynga accounting for 80% of that.2 If they want to keep making that money as users switch to mobile, they have to figure out a way to avoid the 30% tithe Apple or Google would take from app revenues.
Facebook’s current solution to this problem is by betting big on HTML5 to get around the current mobile gatekeepers. So far, there’s been very little heat coming from that announcement. Granted, it’s only been four months since the big initiative was announced, but the paucity of their showcase doesn’t bode well for their HTML5 efforts so far.
Zynga itself seem to back this up, as they’ve been silent about their HTML5 efforts since they released their initial suite of games. Given that they also released all three of those games and more for the iOS App Store how much return are they seeing on HTML5 apps compared to other native apps?
It’s not just Zynga either — for developers, there’s no clear reason yet to develop for Facebook’s mobile ad platform. The viral loops — the douchey notifications and requests that show up in your Facebook feed that are the reason for Facebook Platform’s success — aren’t nearly as strong in mobile as they are on the desktop. And so far most users tend to prefer native apps over web apps.
Errata
I’ve largely ignored their international revenue streams, even though they are significantly underperforming compared to the U.S. and have the most room for growth. If their ad quality was on par with the U.S. that would account for an additional $6 billion in revenue, which would make it a lot easier for them to book Google numbers. It’s a fair argument to say that they’re going to be able to ratchet that up — my question is how easily? Facebook’s international ad quality hasn’t increased as quickly as their domestic quality and doesn’t address the very big mobile problem looming over the horizon.
Of course, Facebook’s got a card up it’s sleeve when it comes to mobile —their long-in-the-works phone. And if Facebook’s going to release it, it’s probably going to be this year. But between Apple, Android, Microsoft and maybe Amazon, it’s not a sure thing for Facebook, even with 845 million users.
Let’s assume for the purposes of this exercise that domestic includes Canada. ↩
That reliance on Zynga, a company that is losing $150 for every new customer they bring in, is another can of worms. ↩
Ina Fried at AllThingsD on Nokia’s earnings report:
As part of Thursday’s earnings report, Nokia noted it received $250 million from Redmond in the first of many quarterly “platform support payments.” It’s part of what the company says will ultimately be billions of dollars in support of its shift to Windows Phone.
Nokia also notes that it pays Microsoft royalties on each phone and has guaranteed minimum commitments, an amount it also expects to ultimately be measured in the billions of dollars.
Hilarious. What’s great is that Microsoft could’ve clearly offered Nokia the OS without royalties and it still wasn’t a sweet enough deal for Nokia. What’s the royalty fee? Apparently around $15.
That means Nokia has to sell over 66 million phones for Microsoft to break even on a billion dollars. On two billion dollars Nokia has to sell over 133 million phones. That’s comparable to the number of iPhones expected to be sold this year. Nokia’s sold a million Microsoft Windows phones so far — a decent beachhead, but they have miles and miles to go.
Dan Rowinski at ReadWriteWeb, on webOS’s chances now that HP has laid out its open source roadmap :
What it may boil down to is this: Palm may have been ahead of its time with webOS, but it fell behind the times when the native app environment exploded. With the coming wave of HTML5 mobile Web apps, the time for webOS to shine may come again.
There’s two premises at the heart of Rowinski’s article. The first is that at some point in the future, web apps won’t suffer from the performance problems and lack of features that distinguish them from native apps. The second premise, supported by the first, is that the ability to write one app and easily port it to other mobile operating systems will somehow make webOS more attractive as a platform.
Let’s put aside the problem that webOS was poorly constructed, doomed from the start and unlikely to have gotten much better as a platform since HP noticed that it was starting to smell. Rowiniski’s argument still doesn’t scan to me.1
The problems with the first premise are pretty clear, and you can look at games for that. A web app version of Infinity Blade may be theoretically possible, but there’s no way we’re going to see one anytime soon. And even if we do, native equivalents will have advanced to the point that audiences wouldn’t care. Think about the HTML5 version of Quake that was released — 14 years after the original. It was neat, but consumers couldn’t care less.
Infinity Blade may be an extreme example, but it’s not just games. The Gmail App for iPhone — which was just a web app wrapped in a thin native client — was a pile of shit. Even if they’d nailed the performance, the Gmail App still didn’t look or feel right.
Here’s a quick thought experiment: Think of any native app that may be indistinguishable from a web app in terms of features and then think about which one you’d rather use. There’s something off about web apps, especially thinly-wrapped ones which reek of pretense. Apple knew this, which is one of the reasons they cleared the path for native apps.
The other issue with the first premise is the lack of access to system APIs for web apps. Applications created using Enyo, the Javascript framework behind webOS apps, will be able to request access to elements like the camera and the accelerometer. But those aren’t the only device elements that matter — not even close. iOS’s Notification Center and Twitter integration are both occluded from web apps. So in order to get around that you have to put a thin wrapper around the web app, but what you’ve actually done is make a subpar native app.
The second premise, about the easy portability of webOS/Enyo apps making it more attractive to developers — and thereby increasing the attractiveness of the webOS platform to a vendor like HTC — makes even less sense to me. If you’re a user on a non-webOS device, you get to choose from all the native apps that would be impossible to create with Enyo and the web apps. If you’re using a webOS device, you just get the web apps. As a developer, you can’t release apps that harness the power of a native development environment. And as an OEM like HTC, your phones are on a platform that won’t have an app like Infinity Blade.2
It’s not that web apps are dead. But the clear future for them right now are very specific use cases that don’t need the features available to a native app. The Financial Times, the only web app I can think of that has had any sort of success is one example of this.3
Here’s another thought experiment: what’s the likelihood of a deeply-troubled company like HP only releasing webOS tablets or phones?
Enyo is likely to survive in some form. I’d hardly say the same about webOS.
I’ll make the disclaimer right now that I’m an amateur when it comes to programming and development. But I did a fair amount of research for this. If I’m wrong — I’m always happy to know. ↩
At least not easily, in the web apps future that Rowinski is envisioning. I’m ignoring the PDK, but Rowinski is writing about web apps. Also, I could find no mention of the PDK on the Enyo site so I have no idea how it factors in to the future of webOS. The “Welcome to webOS Open Source” blog post makes no mention of it or any plans for it in the future. ↩
Lets not forget they were forced to go the web route in order to avoid the 30% slice Apple wanted to take off their subscription revenue. ↩
Galen Gruman, at InfoWorld:
But Apple products, not just the iPhone 4S, took all five top spots in Good’s list of most activated devices: the iPhone 4S at 31 percent, the iPhone 4 at 18 percent, the iPad 2 at 14 percent, the original iPad at 4 percent, and the iPhone 3GS at 3 percent. The rest of the top 10 were Android smartphones: the Samsung Galaxy S II, Motorola Droid Bionic, Motorola Droid 3, HTC Evo 4G, and Motorola Droid X2—each at 2 percent or less.
Apple’s not just beating other devices in the enterprise. They’re humiliating them. How does a phone that just came out a few months ago end up being beat by a tablet? And not just a tablet but a tablet released almost two years ago?
Oh, and someone shoot a BBM to Thorstein “I don’t think that there is some drastic change needed” Heins — Blackberry devices couldn’t crack the top ten.
In their target market.
Actually, that nickname’s too long — but how about “No Change” Heins? Yeah, that’ll do.
Looking through Google’s fourth-quarter results, there was one conspicuous number missing: their mobile revenue run rate. Last quarter, Google said they were expecting $2.5 billion a year from mobile devices — and that’s all mobile devices, including the iPhone. This quarter: nothing.
Can we figure out how much Google is making off Android? Maybe. Google reports in three ways: Google websites revenue, Google Network Members’ websites revenue, and other revenues. I have no idea if mobile revenue is part of “other revenues”, but based on the language Google uses, it’s not a crazy assumption.
Google reported $1.37 billion dollars in “other revenues” for the year. Let’s make the extreme assumption that all that is from Android advertising revenue. This means that if the run rate held true, almost half of all mobile revenue Google made this year is from non-Android devices, and most of those are probably running iOS.
Take these numbers with a huge handful of salt. I’m basing it entirely on the fact that Google talks about “website” revenue, which in my mind doesn’t include apps. Maybe it does. Either way, it’s funny that whenever Google talks about Android, they rarely talk about the money they’re making off it. 700,000 activations a day, 250 million activations total: great, but at the end of the day the only thing that matters is how much cash Google’s got in the bank.
EDIT: I forgot about the fact that Google said that 2/3rds of their mobile search volume comes from iOS devices back in September. Browsing from Android devices probably has increased since then, but it’s entirely likely that the majority of Google’s mobile ad revenue is still coming from iOS devices.
John Gruber at Daring Fireball:
What I’m saying is, if LTE’s current chipset sizes and power requirements are not forcing handset makers to go with these bigger-than-4-inch display form factors, then where are the 3.5-inch display iPhone-sized Android (or Windows) phones that support LTE?
Shopping for smart phones on Verizon’s website seems to support my theory. All their smartphones which support LTE have displays bigger than 4 inches; all (or at least most — I can’t say I looked at every single one) of their smart phones which don’t support LTE have displays smaller than 4 inches.