There are people lined up to buy an iPhone and the darned thing doesn't even go on sale until tomorrow.
There is a lot to be said, both good and bad, about this experience, but given the mail I'm getting, I think it is time to revisit why generation one products generally don't work very well, regardless of who builds them or what markets they target.
I've been doing in-depth testing of the HTC Touch phone, and it is the closest of any of the phones I've seen to being an enterprise-quality iPhone, in case you need to eventually provide a short-term alternative. It hasn't officially entered the U.S. market yet, but you can get it at numerous gray market outlets. I'll give you a quick review this phone, which runs Windows Mobile 6, at the end of this post.
In addition, Microsoft is rumored to be working with Apple to get the iPhone to work with Exchange so, ready or not, it's coming and likely to be on your plate by this time next year, when the third-generation phone comes out.
Why Generation One Hardware Products Suck
With an existing product, there is a base of information in the company with regard to what customers want and what they are willing to pay for it. While companies often lose track of this, or completely screw up the quality of offerings for a variety of reasons, by generation three, the company either has a solid offering or is likely thinking of exiting the segment.
Even if the market is mature, like the cell phone market, if a company isn't actually in it they will make assumptions based on facts that often are false. The good news is, a new company entering a segment might actually challenge conventional thought and come up with something compelling, and the iPhone and the iPod are both good examples of the latter.
But, initially, there are opinions generally generated inside the company with regard to all aspects of a product, from hardware to software to services, and these assumptions are then taken and translated by the developers who craft the product and get it on the test manufacturing lines, where some of the initial build and quality problems are worked out. Prototypes are created and circulated, but they are incredibly expensive and often secret, so they generally can't be put into service by anyone until too late in the production cycle for major changes. We did see some minor changes that came out of this process with the iPhone when they changed the screen and reported better battery life recently.
As the production lines come up, testers inside and outside the company get the products and much of their feedback goes into the generation 2 device, which is often rushed to market to address any glaring customer problems.
Depending on the nature of the problem, the second-generation product will come out rather fast (the new iPhone is supposed to show up in October, for instance). But this device only captures the initial impressions because no one has actually used the product through anything approaching a full-duty cycle.
After the device has been in the market for several months, enough data is captured by the sales and support organizations, Web sites, and technology publications to define the first truly customer-driven changes to the offering. In addition, the lines have been up long enough to determine where the primary hardware quality problems are and recommend changes in physical design to both reduce manufacturing cost and improve product reliability.
This is your generation 3 product, and why many of us typically recommend that folks, if possible, wait until that product for anything that is, based on your budget, really expensive.
Why Doesn't This Apply to Software?
It can; it depends whether the software was designed to be fixed on the fly. If the software is embedded, it generally has to follow a similar path to hardware because it is tied very closely to that hardware, but if it isn't, it can be patched in the field and there are not manufacturing lines to set up, though for major changes, there are beta cycles to ensure a fix isn't a bigger problem than what it fixed.
The first OS X drop had major issues initially; Windows 95 and NT both did as well. But with the Web, Vista is already close to a generation 2 offering and many argued it didn't even need to be cycled thanks to the in-line patching. Going forward, we may need another metric to determine software readiness because the old Service Patch metric appears out of date. For instance, some folks are already deploying Windows Server 2008 (according to the calls I've been getting this week) and it isn't even shipping yet.
Still, with software, you rarely want to be one of the first and you do want to have some test method you can trust to determine if a new product is mature enough for you to deploy. The exception to this is if you can engage the developer of the software early and get them to make changes specifically for you or fund services you otherwise would have to pay a lot for. This is how Microsoft often gets companies to deploy early and because Microsoft covers much of the bill and you often have more say on what actually results, this works for a lot of folks but clearly not everyone.
For open source products under active collaboration, in a way, you are part of the development team and need to make a determination based on testing when the thing is ready. Going forward, this may become a generic way to determine readiness because we likely have all seen vendors break mature products and, like the Windows Server 2008 example, pre-release offerings that were good enough for what we wanted to do.
The Enterprise iPhone: An iPhone Clone You Can Expense
Specifically with regard to the iPhone, the biggest problem is that it's a smartphone that can't legitimately be expensed because it is designed primarily for entertainment and wasn't (initially) integrated with any of the messaging, calendaring or contact solutions businesses typically use. Given that just the service for this phone costs between $720 and $1,200 a year, taking the total cost for the device up to the nose bleed $3K level, you have to admit that's a lot of cash for an individual to fork over. Some may pay over $6K for one and they'll likely need special help.
Yet we expense RIM BlackBerries, Motorola Qs and Palm Treos regularly and, even for small business, being able to use pre-tax dollars effectively reduces the cost from 30 percent to 50 percent. While there likely will be more, right now there is only one iPhone-like device that will fully integrate like these other phones and that is the HTC Touch, currently only sold in volume in Europe and costing $600 in the U.S. through Dynamism. Unlocked, this phone will run Good Technology or connect natively to a current-generation Exchange mail system for full support of calendar, mail and instant messaging. It's based on a generation 6 platform, and if you are already using Microsoft Mobile products, it should get through the review cycle relatively cleanly.
Still, it lacks a keyboard and really isn't the best for messaging. Folks who like BlackBerries and Treos probably won't like it. But for folks who like the iPhone, it is an alternative that may be expensible. It is also relatively unique and it too has an advanced 3D UI, which makes for good eye candy. The big draw is your folks can likely support it and that makes all the difference in the world.
I've actually been really pleased with the HTC Touch, though I'll go back to a phone with a keyboard as soon as I possibly can because writing e-mail on a touch screen still sucks.