Another topic of discussion at the Bathwick session yesterday was how the emerging world of SaaS and Cloud might impact on the future of CIO roles. In a world where technologists are no longer required to purchase what are fully-functioning business services, why have a CIO?

For me, the value in an internal group to manage business systems comes from having the oversight of what is happening across the enterprise, and how what the business does marries to the technology services it uses. Data and process elements of Enterprise Architecture become core competencies, along with procurement and (most crucially) supplier management..

Unfortunately, whilst to a savvy IT person the above should make perfect sense, it doesn't really sell itself to someone outside of our world.

Keeping the CIO role is as much as anything about selling it to the people who make the decisions. That has to happen at many different levels, but for me at the moment if summed up to my business as moving my department from trying to be experts in technology to being the experts in how our company uses technology.

An interesting breakfast briefing run by the Bathwick Group yesterday morning looking at the subject of Cloud Computing.

The topic is suitably amorphous in definition, but much of the conversation focused on Software as a Service (interestingly, few seemed that interested in the utility computing services like Amazon's EC2). One thread that allowed me to get some things to click into place was the impact that commodity computing might have on the way in which IT costs are attributed to consuming departments.

A number of years ago, when I was working at BBC Worldwide, I was involved in the implementation of an activity-based costing (ABC) model for attributing costs to the departments that consumed the services that we provided.

Never has an exercise resulted in so many people hating an IT department. Looking back, there were a few problems: people don't like being told they have to pay for things they previously got "for free"; people don't like to be told to pay for things they haven't requested; and people don't like to pay for things through internal cost charging.

Not only did we take all this flak, but most of it was by proxy for other internal services (particularly finance), where the costs of running things like the ERP system were being passed through as a technology cost.

SaaS (in particular) does seem to offer a way to break some of these blocks, and provide businesses with a greater level of transparency as to the costs of doing business. Whilst introduction of SaaS services in the beginning might highlight costs that were invisible before, they are generally going to be "real" (ie actual money leaving the company) rather than figments of the imagination of a management accountant, shifting pots of overhead from one cost centre to another.

If procured well, SaaS costs are also going to be truly scaleable. One of the biggest headaches associated with the ABC model was that whilst it made IT costs much more visible, it didn't leave business managers with any choices ("Hmm. That networky stuff looks expensive. I'll have a bit less of it, please"... "Erm, you can't."). Turning off users here and there makes no difference to the overall cost of running a service in the 'on premise' world, but it does in a per user per year model.

All of this scaleability and transparency also gives me hope that one day I will also find the promised land of a CFO taking the hit of (SaaS) ERP costs on the chin and being honest to the business about how much the finance services he or she provides actually cost to run, rather than passing the buck to the CIO...

After nine months work, the project to improve our mobile provision in the UK is at last drawing to a close. The final remaining users to be migrated were the handful of iPhone users, for whom the transition to the new BlackBerry devices was one that demonstrated a real sense of loss. Apple undoubtedly can create technology that users become all anthropomorphic about, and the iPhone has had a massive impact on user interface design that the rest of the market is struggling to get anywhere near. Not only that, but Apple are mysteriously able to maintain a cachet around their products even when they have become the mainstream. U2 flogging BlackBerrys isn't, in any way, shape or form, going to make them cool.

So, having said all of that, why have I run a project that has made me reasonably unpopular with some of the key people in the business? Well, if you haven't noticed, there's been this recession thing going on, and moving to a single voice and data mobile platform will save something in the order of three salaries each year. For a company of our size, that’s significant.

I've seen many articles comparing BlackBerry and iPhone as competing platforms for business and they have, exclusively, focused on the relative merits of functionality (usually concluding that BlackBerry is better for email and calendar, the iPhone everything else). But in this present climate, these are in my mind the wrong questions to be asking.

Since RIM (Research in Motion - the manufacturer of BlackBerry) started deploying devices, they've had a proprietary data network which delivered email & other data via the mobile phone operators to the devices. There are a couple of features about BlackBerry data that mean that it is then possible to get low-cost, fixed rate data roaming contracts:

The first is the technical one that the mobile operators would like you to believe (and in which there is some truth). The BlackBerry data network is highly optimised for delivering compressed data effectively and efficiently. Therefore, the total volume of data that a BlackBerry device delivers is much smaller than if you used standard Internet over mobile network (a la iPhone, Windows Mobile devices etc. etc.), and so it is lower cost for the mobile networks to deploy.

The second is the market economics one that the mobile companies would not admit to publicly. Roaming call charges still make up a substantial part of the mobile networks' revenues. Devices like iPhones and Windows Mobile phones have the capability to make Skype-like "Voice over IP" (VoIP) calls, where they use the Internet data network to make phone calls... hence, if you could connect your phone to unlimited Internet data whilst abroad, you could avoid running up big mobile roamed call charges. BlackBerry data networks won't allow VoIP traffic, so roaming data can be offered on a much more generous basis without risking the cannibalisation of roaming call charges.

Apple managed to change the game in domestic markets with the iPhone as it was the device that really ushered in the beginning of unlimited (or near-as-damnit unlimited) data tariffs. But domestic mobile calls are usually uncharged, or just charged as a bundle within monthly service plans, so there was no revenue to lose with VoIP. Inevitably the same will happen at some point with roamed data, but at the moment it appears to big a leap for the mobile operators.

Recent changes in legislation in the EU for roamed call charges are starting to have an effect, and I know that in continental Europe it is now becoming easier to get inter-country roaming data tariffs for iPhone which don't break the bank. My bet, however, will be that non-BlackBerry data roaming won't be cost-effective until at least 2011.

Until that point, if you have a roaming workforce (and for me, two-thirds of the people with voice and data devices regularly travel overseas), BlackBerry is the only cost-effective platform.