Will flat-rate pricing make mobile data take off?

No. It's a nice start, but the operators need to take several other steps as well.

Recently flat-rate pricing for wireless data service has become a big issue in Europe and some other parts of the world. Data service to mobile phones there has often been metered, with users paying by the megabyte. This led to some frightening stories on the Internet of people accidentally ending up with 800-Euro monthly phone bills for browsing too much. Needless to say, this has made many people very cautious about using mobile data.

Recently T-Mobile in Europe offered a flat-rate data service, in which the user pays a single fixed monthly fee for virtually all the data access they want (the limit is about a gigabyte a month, which is a lot for a mobile phone). Then on November 16, Hutchison Whampoa, the owner of the "Three" wireless network in Europe and Asia, announced its own flat-rate plan (more details below).

The Mobile One network in Singapore just cut its unlimited 3G data price by about 2/3, to around $13 a month, in order to compete with fixed broadband services. And on December 1, the CEO of Vodafone went even further, predicting that within a few years we'll have flat-rate billing for all mobile services, including both voice and data.

All of this sudden price activity, especially the announcement by Hutchison, has created a big stir among mobile-focused commentators in Europe. For example, here's a sampling from a mobile discussion board run by Oxford University:

"I'm astonished, frankly.... this is clearly the mobile internet 'done right.'"

"It seems like a seminal event!...Now we have the makings of a new day in our industry.... Has mobile operator 3...discovered the holy grail of the mobile phone industry?"

"When all mobile operators realise they have no choice but to give in too, a torrent of innovation will rush forth I'm sure."

"I strongly applaud this development, and am looking forward to the next stage of competition in 3G/mobile with open gardens and flat data rate packages."

Are they right? Has Hutchison revolutionized the mobile Internet?

I don't think so. Unfortunately, just offering flat-rate pricing is not enough to make mobile data take off. This is one area in which the US mobile phone market has been a leader, believe it or not. The top four mobile operators in the US have offered flat-rate data for years, ranging in price from $15 to $40 a month. Some of them even let you use your mobile phone as a modem, something that Hutchison bans.

The result? Some happy Blackberry and Treo users, but nothing like a mass migration toward mobile data. In fact, the most aggressively priced data service, Sprint's seductively fast 3G network, is rumored to be producing some of the most disappointing subscriber growth.

Flat-rate pricing is a necessary condition for the success of mobile data, but it's not enough. In order for it to take off, the operators have to do some other things as well. I'll discuss those below, but first for context I need to give you an overview of Hutchison's new offer...


Hutchison's X-Series: Open network, all you can eat data

Hutchison Whampoa announced that Three is moving to flat-rate, all you can eat data with a very open business model. The most eye-popping things they promised were unlimited Skype calling over IP, and unlimited instant messaging. Those are both heresies to the operator world. Hutchison's rhetoric was also heretical:

"This charging structure overturns the traditional telephony model of charging per minute, per message, per click, per event and per megabyte."

"What is free to use on the net ought in principle to be free when you use it on the mobile net."

Hutchison made its announcement accompanied by a laundry list of prominent Web brands, including Google, eBay, Yahoo, and MSN, plus the prominent startups Sling Media and Orb. Some of them will be offering specially-configured services bundled with the phones. The first phones offered with Three's new "X-Series" service will be the N73 from Nokia and the w950i from SonyEricsson.

The price of the service is pretty nice. For five pounds a month (about $10), you get 83 hours of Skype a month, 10,000 MSN Messenger messages, and a gigabyte of browsing and e-mail downloads. That's equivalent to unlimited use for most people, unless you're downloading YouTube videos all day.

The Skype and MSN components are potentially frightening to operators, because they compete directly with the voice and SMS services that provide most operator revenue. The overt embrace of those services is, I think, the most radical of Three's changes.

For ten pounds a month ($20), you get everything above plus 80 hours of SlingBox and Orb service, allowing you to use your mobile to play TV shows, MP3s, and other files stored at home.

Although browsing is unlimited, Three notes that not all websites work well on mobile browsers (a polite understatement), and says it blocks access to adult websites. Also, you're not allowed to use your phone as a modem for your laptop computer.


What it means

A change in attitude. I think the most important impact of Hutchison's announcement is not the service itself, but the new agenda it sets for mobile operators. Hutchison was one of the first operators to deploy 3G, and had been an outspoken critic of open Internet access on mobile devices. In 2004, Three COO Gareth Jones said, "People don't want open access, that's not what our customers tell us they want. Anyone in their right mind who tries to do anything on the Internet with a screen that size has to be nuts."

Given that background, the vehemence of Hutchison's new commitment to openness amounts to a declaration of surrender. I think that changes the competitive situation for mobile operators in Europe and Asia. Hutchison is now competing in terms of who can be the most open, rather than who can come up with the most clever bundle of closed services. Assuming that Three doesn't explode or go bankrupt in the next six months, it is putting enormous pressure on the other operators to match or exceed its openness. That change in dynamic makes it much more likely that the mobile Internet will be freed to evolve as quickly as the wired Internet has.

Flat rate is not enough. But just charging a flat rate isn't enough to make mobile data take off. If it were, mobile data would be taking the US by storm.

I think the most important fact about mobile data right now is that we don't know what users will do with it. Hutchison was right several years ago that just blindly transferring the Web to a phone won't please a lot of people. Screen size and the lack of a keyboard and mouse make the mobile browsing experience very different from browsing on a PC. Limitations in coverage, especially for 3G, make the thin apps development model used by web applications much less attractive on a mobile than it is on a constantly-connected PC.

Historically, the software leaders in one computing platform are almost never successful in a new one. When PCs switched from the command-line interface to graphical interfaces, the established software leaders -- Lotus, WordPerfect, Ashton-Tate, and so on -- were almost all swept aside. When the Internet arose, virtually none of the graphical applications leaders were able to make themselves leaders in Web apps. Now that the mobile Web is appearing, it is naive and foolish to think that the desktop Internet leaders will automatically be the leaders in this new space. They don't understand it, and their existing desktop-based businesses are a hindrance to learning.

What needs to be moved to mobile networks isn't just the Web's applications or price structure, it's the Web's open business model. We need to run a huge number of experiments in order to figure out what applications users will want in mobile data. No single company is capable of doing all that work on its own. The only way to make the experiments happen is to set up a vibrant, chaotic ecosystem in which thousands of developers will be free to rapidly try and fail at a huge number of things. It was that sort of random experimentation that permitted leaders like eBay, Amazon, and Google to emerge on the wired Internet while companies like RealNames, AltaVista, and Chemdex were left behind.


Five steps to make mobile data a success

In addition to offering flat-rate data, here are the other steps a mobile operator must take in order to make that mobile data ecosystem work:

1. Provide a consistent architecture that works offline. This is probably the most critical need. Web applications depend on having a constant connection between the user's computer and the Internet. That's not practical for the mobile Web. Even in countries with heavy 2G coverage, there are lots of gaps in the 3G network, and will be for many years. Mobile Web apps need to work like RIM's e-mail client, which stores both the program itself and the user's data locally and then sends the data to the network when a connection is available.

That means just bundling a browser is not enough. The phones will also need software installed on-device that can manage applications and data when the user is offline. That could be an operating system like Windows Mobile or Palm or Symbian, it could be an applications layer like Adobe Apollo or Java, or it could be other software that the web community will create if given the chance. This software layer will need to be consistent across phones, just as HTML is consistent across all browsers.

Three has already failed this test in the first two phones it picked for X-Series. The Nokia N73 runs Symbian OS with Nokis'a S60 software on top of it. The SonyEricsson w950i runs Symbian OS with its UIQ software on top, which is not compatible with S60 software. Guess what – Three just announced that the N73 is available now, but that the w950i is indefinitely delayed. Three didn't give the reason, but I'm not surprised, since all of the client software has to be rewritten to run on the SonyEricsson handset.

I don't think any operator is capable of setting a platform standard on its own, but they should be encouraging the rapid evolution of a standard, by making their phones open to new software (just as open as the PC is), by offering to help deploy new tools, and by providing free testing to make sure they work on the wireless network.

2. Kill security certificates. The line between websites and applications is blurring, as Web 2.0 architectures allow much more processing to be done on the client device rather than a server someplace in Idaho. Google is pretty much a traditional website, but Google Maps is a hybrid, and Skype is a full PC application that talks to the Web. In the future it will be impossible for a user to tell exactly where an application ends and the Internet begins.

But today the operators treat websites and applications completely differently. The new flat-rate data plans let you browse just about any website you want. But operators are starting to insist that applications obtain a security certificate before they can be installed. The certification process is slow, inconvenient, and unreasonably expensive for small software companies and those that create a lot of applications. Since small software companies are the most innovative, this has an enormous chilling effect on mobile innovation.

This approach is also completely inconsistent with the way the Web works. Can you picture a website paying for certification before it can run on your browser? How many sites would bother? If the operators insist on certificates, they will make the mobile Web a small and uninteresting subset of the real Web, permanently. Certificates have to go.

3. Unlock the user's data. This is the other security-related problem area. Many operators make it very difficult for an application to access the user's data stored on the device, such as the address book, the dialer, and the user's current location. But many of the most interesting new mobile applications need to be able to work with this information. Users should be informed when they give an application access to this information, but it should be very easy for them to say yes.

Like the previous point, this one is scary to the operators, because they worry that a rogue application could make thousands of phone calls or send huge numbers of premium messages, building up a huge bill.

Although that's a legitimate fear, strangling mobile data is a self-defeating solution. The operators will need to adopt the same security model used on the Web -- let the user do what they want, and defend the device via security software.

This isn't as dangerous as the operators fear it will be. We've used the same model for more than a decade on the fixed line phone and data networks, and millions of fully open Palm OS, Symbian, and Windows Mobile devices are on wireless networks today. Despite this, no one has taken down either the wired or wireless networks.

4. Make it easy to discover new content and services. The mobile data ecosystem will evolve faster if it's easy for users to find new services and applications. Today the content discovery tools and software stores on mobile devices, if they are installed at all, are often buried under several layers of icons, or are very hard to use. We need the mobile equivalent of an Amazon.com -- an online content store that's easy to find, browse and search, and that makes suggestions to you based on what you've used in the past.

5. Get ready to go to a flat rate for everything. Vodafone's comment shows that they understand this: the logical outcome of putting the open web on a mobile device is that voice and data merge under a single flat fee. If a Skype call is free, then eventually all calls need to be free, or the users will just switch everything to Skype. Same thing for SMS messages once they're directly in conflict with instant messaging. The operators' old financial model won't evaporate overnight, but it's now officially dying. And it's in Three's interest to move to the new model as quickly as possible -- the sooner it adopts the new model, the sooner it can cannibalize the customers of other operators. I think the race is now on for full flat-rate mobile pricing in Europe. Because the operators there are more competitive than the ones in the US, I think it's very possible that Europe's pricing evolution will move faster than in the US. But the same principle applies everywhere in the world: the operator that moves to the new model fastest stands to gain the most customers.


Let's hear it for desperate operators

Some people have been asking why Three chose to make such a radical change. It may be genuine vision, or it may just be desperation. Press reports say that Hutchison had been hoping to spin out the Three networks, but had to cancel the plans because Three is not making money. If that's the case, Hutchison may have decided that it might as well take a chance on aggressive new pricing.

Unfortunately, if Three is counting on the new pricing alone to bail it out, then I think it'll be disappointed. It needs to take all the steps I've outlined above. But I like the idea of desperate operators being willing to experiment. If we get enough of them desperate, someone will probably eventually make all the right changes. I like to believe it's just a matter of time.