Zigurd's iPad Scorecard

Form factor [8.5]
It's a Web pad that meets expectations for Apple's industrial design reputation. The most notable thing is that, while it isn't the first one, Apple is among the pioneers, and not an aggressive follower. Apple beat all of the major OEMs to this market.

Price [8.5]
Apple priced iPad low enough to remove price as an objection, and even low enough to make potential competitors re-evaluate their Web/media-pad plans. Apple didn't leave a lot of room under their pricing for less-amazing products. Apple took the netbook price challenge seriously and has  a serious response in iPad.

Applications [9.8]
The app store ships about 400 new applications per day. Read that again. Now look at the top selling programming language books on Amazon. iPod Touch has evolved into a game and social media platform, as well as being the continuation of the iPod product line, and iPad inherits all of this. Apple announced two new application/content categories: e-books (and periodicals), and iWork. And it is the best Web pad yet devised.

Software [7.5]
It's a big iPhone. The software appears to be very well executed, and enables iWork and other new key applications. But it isn't a breakthrough. Not even a big “Wow” factor. There was a chance Apple would send Google's Android team back to the labs, but no put-away shots here.

Hardware [9.5]
People who have seen it firsthand all remark on the speed. Apple has entered the ARM CPU business at the front of the class. Success here is crucial to Apple beating netbook economics. However, Apple probably does not dominate this business, and you can bet that would-be competitors are looking at Qualcomm Snapdragon and other fast ARM CPU spec sheets this morning. Apple may find themselves at the front of a very competitive race.

A lot of software effort had to have gone into fully exploiting the GPU, and into other unseen system software, leaving less time for visible UI improvements.

Did Intel make the right move selling off XScale/StrongARM and going with Atom? What will it cost to make Atom competitive and keep it from cannibalizing high-end CPUs?

Ecosystem [10.0]
Apple has created and continues to exploit the best ecosystem in the business. When will Apple's would-be competitors, Nokia and Microsoft among them, learn they cannot succeed by copying a subset of Apple's ecosystem? Google has built their own very powerful ecosystem and can make their own rules. Everyone else needs to take this more seriously or face continued failed market entries. Legacy category dominance is not an ecosystem.

Reviving Windows Mobile: A Bigger Challenge Than Most People Think

Big companies fail at competitive analysis by making the challenges they face fit the framework of a normal project. Reviving Window Mobile is anything but a normal project. iPhone dominates the new generation of smartphones and Android has knocked Windows Mobile aside at many the top tier OEMs. Windows CE is not competitive with Linux as a mobile embedded OS, and the Window Mobile userland is stuck in the C++ era, despite having the C#/NETCF runtime available for years before Android was even an idea.

Colin Gibbs at GigaOm set out to list the ways Windows Mobile could get back in the game: http://gigaom.com/2010/01/13/how-microsoft-can-get-back-in-the-mobile-game/ But, for the reasons I listed above, it's going to take a lot more that . Let's look at the suggestions in the GigaOm article and see just how much more:

Make Windows Mobile free to manufacturers

Right off, we know this isn't enough. Google is giving carriers a slice of search revenue. Bing doesn't have enough revenue to share. Microsoft has to come up with a strategy to bring superior value to the carrier and OEM partners. Gibbs also says “Making WinMo free — but not open source...” But that isn't going to cut it with developers. There is no reason not to make Windows Mobile open source.

Acquire (or adopt) another operating system and ditch WinMo

Like Xerox, Bell Labs, IBM and other companies well-stocked with very smart people, but chronically unable to commercialize their output, Microsoft has more operating systems in their labs than it knows what to do with. More focus, and making the product management side of the business more compatible with the research groups would make some of those operating systems productizable. But if they start today, it will take two years to bring such a product to market. Microsoft better have a plan for the Windows Mobile kernel well under way or they won't have a result in a commercially relevant time.

Build a top-notch app store designed for business users

This will only make a difference if the apps are all C# apps for an upgraded runtime. And even such a remarkable change in the userland won't matter if Microsoft keeps the price of Visual Studio at $799 for mobile application developers. For that kind of money, buying a Mac to make iPhone software no longer seems like such a big hurdle. And the price to get into Android software development is $0.00.

Make Windows Mobile 7.0 a worthy competitor with a focus on the enterprise

That's a good suggestion, but it's a small suggestion. Better to aim at clearly identifiable value for the enterprise user: A secure connection to Office Communications Server, including for IP voice. A VPN in every handset. Data and voice security are serious, high-value issues. Microsoft can, and must in order to demonstrate high and unique value in Windows Mobile, make a big move in mobile data and voice security and in enterprise communications integration.

As Colin Gibbs says in the GigaOm article: “As we’ve said before, it may simply be too late for Windows Mobile to re-emerge as anything but a niche play for a small number of business users.” And that's what will happen if Microsoft views this as a normal project in a normal competitive environment. This isn't a new version of SQL Server. This determines whether Microsoft is in or out of mobile.

In addition to the above, Microsoft should be doing more:

Make the userland pure C#

Ditch the legacy applications. They don't matter. Google and Apple started from zero with applications. You can too. And in the end you will have to.

Use the power of OCS

Office Communications Server is revolutionizing corporate messaging and voice communication. Put it in every mobile handset. Provide a SAAS version for smaller businesses.

Don't let Apple take the handheld games business without a fight

You have in-house titles nobody but Sony can compare with. Why aren't these on your phones? Might I suggest Mobile Halo Wars?

Zune for media

Cannibalize Zune to add value in mobile. Every phone should be as good as Zune at playing media.

Buy Opera

If you can't make a better browser than Opera, buy Opera.

Don't complain about how much it will cost

How much did you spend on Danger? No excuses.

Is the MagicJack Femtocell For Real?

The answer is, probably, yes.

The most interesting announcement I saw in CES coverage was the MagicJack femtocell. It's one of those products that is brilliant technically, and a brilliant and audacious business idea.

What the MagicJack femtocell does is: It enables your mobile phone to connect to the MagicJack VoIP phone service as if it were connecting to your mobile service provider's femotocell. This is far better than using a SIP client application on a smartphone. First, you don't need a smartphone, which is incompatible with the budgets of many of MagicJack's customers. And it just works. Your phone will display a network name indicating you are connected to MagicJack and that's it. No added software in the phone.

Well, that sounds simple enough, but it is enormously disruptive: MagicJack just jacked your mobile carrier's subsidized handset, and a slice of GSM bandwidth, and used them to to connect to their $20 per year VoIP phone service. And it should work as smoothly as using your carrier's femtocell. And it is cheaper than your carrier's femtocell.

How is this all possible? Once you start to look at it, it begins to look unlikely:

1. Femtocells are not cheap. A 3G radio. A GPS receiver. A router. Considerable amounts of software. Integration with your mobile carrier's operating software (so you don't stray into areas where they do not own bandwidth licenses). Etc.
2. Femtocells take big-company resources to develop: Huawei and Alcatel Lucent, to name a couple early entrants to the market, spent tens of millions each on femtocell development. MagicJack's total revenues amount to tens of millions.
3. Femtocells are not easy: How did MagicJack take a femtocell, make all the modifications needed to connect it to their VoIP network, test it, and ship it? MagicJack is a relatively large VoIP operator, but companies like MagicJack, which has been slammed by some for having clunky, adware-filled software for their VoIP dongle don't suddenly become past master at femtocells, ripping out their MSC-connected software guts and replacing them with what can be described as a Very Fine Hack.

And yet, that is what they claim to be demonstrating.

If they are doing this, it has to be made of off-the-shelf parts. Integrating a system like what they claim to have is hard enough. Let's put aside the idea that they did school everyone at what you can do with femotocell firmware and see if there is a path of least resistance to what they claim to have.

First, the femtocell itself: I believe MagicJack must be using a 2G femtocell. This makes sense because they have no need to be your mobile ISP. No requirement for 3G data. This is for VoIP calls, and, if you are home, you, and your smartphone, if you have one, can connect to your WiFi network.

Second, the software: MagicJack's requirements seem unique. Who would make something that would connect a femtocell to a VoIP system? The answer to that question, as with many fine cheap telephony hacks, is in developing markets. Who would substitute an Asterisk server for a proper MSC? A rural telco in the developing world is who. And here is one supplier of just such a system: http://www.go2callsoftware.com In fact, Go2Call Software seems to be a perfect fit for what MagicJack is doing: Femtocell to VoIP.

Does this make it any less amazing? Not really. It is still one of those brilliant and audacious ideas you find you are rooting for. It takes cross-functional thinking – skillful “intellectual trespassing” - to figure out something as, well, devious, as this.

Hardware, check. Software, check. Now for the audacity of it all. Is it kosher, or even remotely legal, to operate a 2G femtocell without a network operator's permission? Is it legal to sell femtocells if you don't own spectrum licenses? This is the part that remains to be seen. But I hope it is. Disruption and change bring opportunity.