The big news this week for Apple wasn't the new 3G iPhone. It was the business model behind the next-generation iPhone, and the threat it poses to Research in Motion (RIM). Apple's model depends on developers. RIM's model depends on devices.
If history repeats itself, the developers will win. Just ask Microsoft.
More on that in a minute. For now, consider the superior TCO (total cost of ownership) argument that Apple now has for both developers and end-users. Many enterprises are going to find the cost/benefit analysis of RIM vs. iPhone favoring the iPhone. RIM's BlackBerry Enterprise Server (BES) solution costs $5 to 10 per mailbox per month (for Exchange), plus an additional $10 per mailbox per month for BES, which includes a combination of licensing plus the cost of administering BES. Not so cheap.
The iPhone? It's still going to cost $5 to 10 per mailbox per month (for Exchange or Zimbra or whatever your mail service happens to be), but the extra $10 charge is wiped away. Gone. This leaves the enterprise with a two-times price advantage for the SaaS/iPhone world, which doesn't even include the cost of the device, which also continues to plummet.
Again, RIM's business model depends on extracting maximum value from each device/user, and it does so to good effect. Apple's business model is shifting to be about ubiquity of devices, and then the monetization of the applications.
Which will be better? Well, that depends on how one feels about developers and their impact on markets.
Microsoft's success stems, in large part, from cultivating a worship of developers. Back when Microsoft was firmly grounded as a platform company, it drove the vast majority of its revenue through partners. The company has lost its way a bit in the past few years with regard to developers, but one can still attribute its long-term success to a focus on developers/partners.
Let's go back to Zimbra, given that it stands to benefit enormously from a switch to the iPhone, even despite its support for the BlackBerry.
Zimbra supports a wide range of devices (Windows Mobile, Symbian, BlackBerry, Palm, iPhone, etc.). Now, while Zimbra will support iPhone 2.0 as soon as Apple releases its 2.0 iPhone software, the company is mobile device agnostic. The company prefers standards-based protocols, but works with a variety of phones.
Even so, Apple is going to create real challenges for the RIM business model because it makes RIM's insensitivity to developers so glaringly obvious. For a company like Zimbra (or Oracle/Alfresco/Name your favorite ISV), developers have a choice: Support BES (which requires an upfront and significant cash outlay on their parts, a new Windows-based server, partnership, etc.), or support the iPhone. Requirement for iPhone support? A free SDK and an iPhone. That's it.
Now, it's likely that Zimbra et al. will support both. This is business, after all: Corporations exist to make money, and will do whatever is legal and necessary to do so. But I'm betting that iPhone support will happen first to an increasing degree. That's not good for RIM.
It's also not good for RIM just how cumbersome and costly it is to set up employees with BES support. Many will chafe at the burden of introducing a new Windows-based BES server with a bunch of new licenses and administration required. Companies will discover that the iPhone may well be cheaper, easier to administer, and offers a comparable feature set through ActiveSync plus the iPhone.
There's a duopoly of Exchange and RIM today, but the iPhone promises to crack that wide open. With the iPhone, we can finally have choice in messaging servers and an awesome new enterprise-class device. Now we just need Apple to use an open standard for sync...but that's another blog post.