The (partial) state of the mobile data market

Despite all the hype and excitement about the mobile data market, it is very difficult to get reliable data on how it's actually developing. The mobile operators don't like to release full details of their sales, and surveys of users cost a lot of money to conduct and therefore are usually available only to people who pay.

We're left to chew on anecdotes, partial information released by companies that are trying to push a point of view, and unscientific "polls" of online enthusiasts.

So I'm always on the lookout for more rigorous information. Recently I came across several fairly good sources of data, and they give some interesting perspectives on what's happening with mobile data. It is by no means a complete view, and most of it is US only. But I think it's worth sharing.

Steady, unspectacular growth

The overall picture of mobile data is one of steady but unspectacular growth. It's a bit like watching a tree grow -- you can't see anything changing day to day, but if you walk away and come back in six months you'll notice the difference. SMS continues to be the dominant service, especially in Europe, and there's no sign of some other service surpassing it.

Is the growth rate good or bad? It all depends on how much growth you were expecting, and how fast you wanted it to happen. The one thing I think is very clear is that each country market is different, and you can't classify any of them as leaders and laggards. They're just unique.

Here are the details:

Capabilities of mobile phones in the US

The Pew Internet organization has been surveying Americans on their Internet usage for years. A couple of the questions in their survey ask about the data capabilities of their mobile phones. In the most recent results, from early 2006, 75% of mobile phone users in the US said their phones are capable of texting. 63% said they can play games, and 39% said they have cameraphones. Here's the full chart:

Nothing there stands out as shocking, although I expected the penetration of cameraphones to be higher. My guess is it has gone up in the last year.

(Note that some people could have capabilities in their phones and not realize it. So the question tells us as much about awareness of features as it does about the phones themselves.)

What people do with their phones: Nothing else rivals SMS

The chart below examines the percent of mobile phone users in the US and several European countries who have ever performed various tasks with their mobile phones. The source is M:Metrics, Q4 2006, and the numbers were quoted in a presentation by Orange / France Telecom.

The figures show that there's no other mobile data service with near the penetration of short messaging service (texting). That's not really news, but it's striking to see the hard numbers. About 80-85% of people in most of the big European countries have ever sent a text message, with France lagging slightly (at about 75%). In the US, almost 40% of mobile phone users report that they have sent a text message.

The next closest service is picture messaging, with 20-30% of mobile users in the big European countries saying they have received photo messages at least once. In the US, the figure is 15%. It's ironic that photo messaging is in second place, since it's generally considered a major disappointment. What does that say about the other services? Well, none of them generally crack 10% usage.

Is the US really a laggard? The other thing in the chart that really stood out to me was that the adoption "lag" of US mobile users varies depending on service. The US is far behind in SMS, MMS, and playing music on the phone (the last one is, I'm sure, due to the strength of the iPod in the US). But in the other categories, the US is in the middle of the pack, or even ahead (somebody explain the ringtone result to me, please).

It's always fun to stereotype the US market as primitive in all areas of wireless, but the adoption numbers don't support that. It just looks different.

What does it all mean? Orange's spin was that it means we're just getting started in mobile data, and everyone should wait patiently for the good services to take off. They showed the following growth projection from Ericsson as evidence:

No offense to Orange, but that is basically a statement of faith rather than analysis. If you're a cynic, you'll point out that the chart assumes compound growth will continue uninterrupted for a decade, something that is often true for technology specs but is rarely true for technology markets.

What we really need is time-series data, so we can see what's growing and what isn't. Unfortunately, Orange didn't present any numbers like that, but the research firm Telephia did, in a separate presentation. Unfortunately, their numbers were US only, and they didn't cut the usage categories in the same way as France Telecom. But they still show some interesting trends...

Mobile data growth in the US

Telephia measures mobile data usage by analyzing the monthly bills of mobile phone users. This should give very accurate information on revenue and number of users, but it doesn't track physical usage. Because some services are billed per-use and some have monthly subscription fees, it's hard to tell how heavily people are using the services listed below.

Telephia reports that billings are growing steadily for a wide range of mobile data services. The chart below shows total US operator revenue for mobile data from Q3 2006 to Q1 2007. (These figures include anything that passes through the user's phone bill. Applications and services paid for separately by the user are not included.)

The chart is in billions of dollars, so it shows that in Q1 2007, total on-deck US data revenue was about $4.6 billion. Is that a big number or a small one? Well, total service revenue for the US mobile operators is about $32.5 billion per quarter, according to the CTIA. So mobile data is about 14% of mobile billings.

Where is e-mail? I can't find e-mail anywhere on the chart. I'm very surprised they didn't break it out separately.

Strangely consistent growth rates. The weirdest thing about the chart is that everything's growing at the same rate. In the real world, that sort of thing doesn't often happen. I wonder if a lot of the growth might be driven by people buying service bundles, where they pay a flat extra rate per month to activate a bunch of different services, and then the revenue gets allocated across the services by the operator. That would cause everything to grow in lockstep.

If that's what's going on, then these numbers really might not say much about usage -- what they'd be tracking is the ability of the operators to sell services bundles.

Anyway, the numbers show that the US operators are making pretty good revenue from mobile data. I didn't make a chart of this, but in general, the growth in mobile data billings is large enough to make up for the ongoing decline in mobile voice revenue. So the operators aren't getting rich, but data is helping to keep them from getting poor.

More details. Telephia lumps a lot of different things in the "Downloads" category. For Q1 2007, they gave more details on that category. So I can't give you a time series, but here's a more fine-cut look at how mobile data revenue looked in the US at the start of 2007:

Premium SMS is mostly ringtones paid for via SMS, plus voting for things like American Idol. Audio is downloading and streaming of songs. The other categories are self-explanatory. I feel bad about the tiny size of the applications category, but keep in mind that most smartphone apps are sold through the web and then synced onto the device, and so don't show up in operator billings.

Number of users per service. Telephia also reported the total number of users for each service. As we saw in the Orange chart, SMS has the most users in the US (although the gap between it and the other services isn't as large as in Europe).

Revenue per user. Combining the user and revenue data, we can estimate monthly billings per user for each service:

You can see why the operators like premium SMS. And look at WAP! It never lived up to the original hype that it would become the mobile version of the Web, but as a tool for getting things like sports scores and weather reports, it's not doing too bad. (Whether it's paying for all the money that was invested in it is another story.) Video's generating the most revenue per user, but with a very tiny user base. Audio revenue (which is revenue from listening to songs, not ringtones) is fairly close to what Apple gets from iTunes users (the average iTunes user downloads about 3.3 songs per month, or about $3.30) (link).

Usage doesn't follow capability. And now for the "big" mashup. We can combine Pew Internet's figures on phone capabilities with Telephia's numbers on service usage to figure out roughly what percent of US mobile customers who know they have a given feature on the phone ever actually use it. The results are interesting:

For communication-related services, the percentage of users is quite high (although remember that we don't know how heavily the features are being used). But most mobile users are not adopting the entertainment features in their phones. That's exactly what you'd expect if only a limited percentage of the population were interested in using their phones for entertainment, which is what a lot of user studies have shown (link).

The lesson: If you're an operator or handset vendor, be careful about pushing phones that are a kitchen-sink collection of expensive features. The odds are very good that you'll spend a lot of subsidy money on people who won't ever adopt the underlying services that were supposed to justify the subsidy. It's much better to offer a variety of phones specialized for different types of user, and let them pick the ones they want.


As I said at the start, it's an interesting collection of tidbits, but far too US-centric. If you live outside the US and have information to add on your market, please post a comment.

Total revenue of the operators: link
Orange's presentation at the Global Mobility Roundtable: link
Telephia's presentation at the GMR: link
Pew Internet: link

A new way to measure the popularity of the iPhone

True story: Back when I was working at Apple, we received a report that there had been a burglary at a company that had a lot of Macintosh computers. The thieves took every one of the Macs -- and left behind all the Windows PCs.

We were delighted. Not about the burglary, mind you. I'm afraid we were pretty callous about that, and besides we figured the company was insured. But we loved the idea that even criminal lowlifes knew which was the better computer.

In fact, we loved that idea so much that we made a commercial about it. It was cool -- deep blue lighting, lots of shadows, and two silhouettes moving silently through the office. One of them starts to pick up a PC, and his companion hisses, "not those, just take these." And he points at a Mac.

The Computer for the Rest of Us, indeed.

At the last minute, somebody in the company realized that the commercial might just possibly be viewed as an endorsement of crime, besides which it made us look like complete mercenaries. So I don't think the commercial was ever aired. We just watched it a lot at communication meetings.

I thought of that old commercial today when I was in an AT&T (formerly Cingular) mobile phone store. There was a big video display for the iPhone, but no phones on the counter. "Did they sell out?" I asked the clerk.

"Nope. Someone broke in and stole them," he said. "We saw it on the surveillance video. They were in and out in 30 seconds. They ran over to the iPhones, cut the cables, and left."

"Did they take anything else?" I asked, thinking I saw a new commercial in the making.

"Oh yeah, they took the Treos too."

Sorry, Apple. No commercial.

And sorry, RIM. They left the Blackberries. Personally, I would have grabbed a Pearl on the way out.

By the way, if you're thinking of buying an iPhone on eBay, keep in mind -- the other thing the AT&T guy told me is that the iPhones are all serialized, and Apple can permanently turn off the ones that were stolen.

Impact of Amazon Flexible Payments Service: Computing as a utility

The announcement of Amazon FPS made my whole week, on a lot of different levels. I'm excited about the service itself, I'm excited about what it means for the development of web applications, and I'm excited about what it'll eventually do for the mobile data world.

Okay, I'm just excited.

About FPS. Before I talk about what it means, I should give a quick overview of what it is. FPS is a web service, meaning it's a set of online APIs that the creator of a website or web application can use to perform tasks. What FPS does for you is billing -- you can use it to accept payments for something you sell online. Basically, you transmit the customer's info to Amazon, and they take care of the credit check, credit card processing, billing, and so on. They send you the money, less a percentage cut that they take.

That's not at all revolutionary. PayPal and Google Checkout offer the same thing already. Amazon's cut is about the same as PayPal -- about 2% to 3% of your revenue, depending on the amount of business you do, plus 30 cents per transaction. Google is a tad cheaper, plus you get AdSense credits for using it.

(For more information on FPS, there are good articles here and here).

What impressed me about FPS is its flexibility. Amazon says you can set different payment terms for every customer, set up subscriptions and multiple payment schedules, manage a store in which you pass payments from a customer to your suppliers, set up either pre- or post-payment systems, and most importantly you can manage micropayments down to a couple of pennies per transactions (link).

The competing systems either don't offer this at all, or do it badly. I think FPS is a really important change to the competitive situation in payment services. And, because the payment services are all available to any website, that means it's an important change to the whole web platform.

New forms of online business. So far, e-commerce online has been limited mostly to selling things that we could already get through regular stores -- books, clothing, software, etc. One of the main culprits for this was payments. The current credit card system, with its strong discouragement of small transactions, makes it very hard to sell anything priced below a few dollars online. I think the most interesting use of online commerce will be the creation of markets for things that we can't buy through stores today. Most of those things are intellectual property of various sorts, and the natural market for them is a buck or less a copy. So the payment system is a big barrier.

I won't recap my whole argument for minipayments; I wrote about it recently, and you can read it here. Minipayments have already changed the world in music, where Apple's proprietary minipayment system in iTunes has revived the market for music singles, something that was virtually dead in stores. Another example: iStockPhoto has created a market for low-cost stock photography. By creating an easy system of practical minipayments, Amazon FPS will help to enable the creation of lots of iTunes and iStockPhoto equivalents for other products and forms of intellectual property. Think short stories, art, games, and probably a lot of other things we haven't even thought of yet.

I know FPS isn't perfect -- for example, small payments have to be aggregated and then billed in a single larger transaction. But it advances the state of the art dramatically, and more importantly it challenges Google and PayPal to improve their own minipayment handling. That competitive dynamic should eventually result in a truly great minipayment mechanism online, no matter who makes it.

Amazon vs. Google: A contrast in strategies. I think Amazon's approach to web services makes Google look bad. Both companies are taking on PayPal, but Google's approach so far has been pure blunt force -- duplicate PayPal's features, underprice them a bit, and tie it to another Google product (you get AdSense credits for using Google Checkout). Let's compete by duplicating someone else's features, underpricing, and tying back to your dominant product. Does that remind you of a certain company in Redmond?

In contrast, Amazon has been trying to find holes in the infrastructure that nobody has filled yet. Its storage and compute services provided very important infrastructure that helped accelerate the growth of Web 2.0 companies. Although its payment system is not as unique, the emphasis on minipayments is, and I think it too will play an important part in the online ecosystem.

Bottom line: Google is often copying, Amazon innovating. I'd say that I'm disappointed in Google, but actually given their size they would crush everyone else if they were also innovative. So maybe we should be grateful.

What will Amazon do next? Their pattern is clear -- they're picking out things that they know how to do well (because of their retail operation) and turning them into services for other developers. A logical next step would be if they offered developers the infrastructure needed to set up an online store -- order tracking, support request tracking, inventory, displaying merchandise, etc. That would work with their other services, and would put them in a position to start draining business from eBay.

I'd also love to see them offer some sort of unified product and content discovery system. One of the things missing from the online ecosystem is an easy way to find goods and services that are for sale online, and comparison shop between them. You can use search for it, but it's not very well organized, and comparisons are difficult. eBay kind of does that, but you have to be registered as one of their sellers, and eBay does the billing. I'd love to see a looser directory than eBay that doesn't take the payments directly, but just points you to things you can buy.

That's what I thought Google Base would evolve into, but Google hasn't made the move yet, so there's still time for Amazon to seize that territory.

What it means for mobile. You can probably guess what I'm going to say here. The operators consistently charge up to about 50% of revenue for any songs, games, or other content sold through their networks. The mobile software stores like Motricity and Handango charge about the same. Amazon, Google, and PayPal each take about 2-3% of revenue, and that cost is likely to decline due to competition. As the wireless Internet takes hold, how many users will be willing to pay 50% extra just for the pleasure of having a game appear on their Sprint or Verizon bill rather than their Amazon bill?

If an operator bit the bullet now and priced competitively, they might be able to hold onto about 10% of revenue in exchange for the greater convenience of running content purchases through the mobile bill. But a 50% cut is like waving a red flag in front of a bull. There's no way Amazon and friends will be able to resist the temptation to target the mobile web. The question is not if, it's when.

The name of the game is infrastructure. In an open, decentralized computing environment like the Web, the best way for a software company to succeed is to create a control point -- to offer a piece of critical infrastructure that others need, and build a franchise around it.

Google understood that concept with search + advertising, and did well with maps, but has been remarkably inept at creating other strong points. I think that's because, to be blunt, engineering PhDs don't necessarily make the best business strategists. Google, if you want to go to the next level, ya got to hire business people who are as smart as your technical people. And you have to give them some authority.

Microsoft seems to get it, but is still trying to retrofit its applications into services rather than really thinking through what's needed in an online ecosystem. Apple seems to understand, but so far hasn't been interested in opening up its services to others (it could easily have turned iTunes into a content discovery and billing service, long before either Google or Amazon hit the market). Some other big Internet companies, like Yahoo, don't seem to really understand yet that this is the competitive battleground of their future.

Amazon is the one major web company that seems to both understand the situation, and be able to consistently come up with good new services. They already have two strong points (computing services and storage), and payments looks to be the third. If some of the other players don't wake up soon, Amazon's going to end up in an extremely powerful position online.

Why I work in the tech industry

Because of the cool toys, of course. But for me, there's also something else: Making a difference in peoples' lives.

I'm sure that sounds corny, and not everyone feels the way I do, but for me there's something very compelling about the ability of technology done right to improve the lives of lots of people, in ways you don't expect and can't plan for. I got my first taste of that years ago, when I was a Mac software developer in the early days of desktop publishing. I was mostly creating PostScript fonts, which is a thoroughly commoditized process today but was difficult back then, before third party font-creation products were available.

There was a lot of demand for fonts, so I would go to Mac trade shows and sell the fonts from a booth, $30 or so a pop. I didn't get rich, but it was fun and I learned a lot. One of the things I learned was that you have to take the work really seriously, even on something as trivial as fonts. A man walked up to me at one of the shows. He didn't look like one of the technophiles there, just a normal guy in his 30s, someone you might run into at the gas station. He spoke with a slight southern accent.

"I like your fonts," he said, and smiled.

"Thanks!" I said. One of the coolest things about those shows was that you got to talk to customers.

"Especially that one." He pointed to a font that imitated calligraphy, lots of curves and soft angles. "When my niece died, we used it to make the engraving for the tombstone."

I thanked him and expressed my condolences, as politely and gently as I could. But I was in shock the whole time. I thought my fonts were for throwaways, things like flyers or newsletters that people glanced at and then tossed. It was fun to make them, and a good learning experience, but I wasn't expecting anyone to use them for anything serious. Certainly not as serious as saying goodbye to a loved one who died young.

What it taught me was that making a new enabling technology is a trust. You don't know what people might do with it in their lives, and so you should always take it very seriously and make sure it really works the way you promise it will. Because chances are, in a world this big, someone will depend on it an awful lot.

That lesson came back to me a few weeks ago, when I saw a photo essay in the New York Times showing real-life people with the avatars they use to represent themselves in computer games (link; you can read about the underlying book and exhibition here). There were the examples I expected -- the chunky geek guy who shows himself as a red-haired ninja girl, etc. But most of the avatars looked a lot like the people who made them -- thinner maybe, or with better hair or bigger, uh, appendages. But quite recognizable.

And then there was the photo of Jason Rowe, a young man in his 30s who has muscular dystrophy (link). He's in a wheelchair, with a breathing mask on, and a report by NPR said he can move his hands just enough to control a character in Star Wars Galaxies, which he plays about 80 hours a week (link).

His avatar is a huge husky guy in armor, waving politely at the viewer -- something that you imagine might be difficult for the real-life Mr. Rowe. In all of the other pictures, you get the impression that the real person is saying to you, "here I am in real life, the avatar is my mask." But Rowe's avatar seems to say, "this is the real me over here where I can move around, don't judge me by my physical shell."

I've never met Jason, and I may be reading way too much into his picture. But if he's not thinking that, I know there are a lot of other people on the Internet who are. I doubt the folks who wrote Galaxies or Warcraft or Second Life were expecting to have this sort of impact on peoples' lives, but that's what happens when you work in tech.

And Jason, thanks for posing for that picture. You reminded me what this industry is really about.