The rise and rise of cloud has been dominating the headlines for the past few years, and for CIOs, it has become a more serious priority only recently. People like cloud computing. Well – at least they like the concept of cloud computing. It is fast to implement, affordable, and scales to business requirements easily. On closer inspection, cloud poses many challenges for organizations. For CIOs there are the considerable challenges around how you restructure your IT department and IT services to cope with the new demands that cloud computing will place on your business – and often these demands come from the business, as they start to get the idea that they can get so many more business cases over the line for new capabilities, products and/or services, as they realize that cloud computing lowers the costs and hastens the time to value.

Some people see this model and focus on how great it is. I see cloud computing, and it has me thinking about how poor the traditional sourcing models have been. However, the reality is that many organizations will not move everything to the cloud – and some will move nothing. Compliance issues, security, lack of visibility, additional overhead required, and a potential increased risk will see some organizations shy away from cloud – and at the same time miss out on the benefits that it could bring. So this got me thinking – what if you could get some or all of the benefits of cloud computing through an internal solution? Yes – there is virtualization and the "private cloud" – but that model still has its issues. IT departments still have to buy the equipment, virtualize it, integrate the apps and the data through the middleware, and then manage the whole solution. And what I have described there is the real challenge with IT these days – bringing everything together. Yes, it is nice if you can outsource some servers, storage, and even applications to the cloud, as this lowers costs, shortens time-to-market, and means less management. But you still have to integrate the solutions across a business process – and that process is one that constantly changes.

I therefore propose a new model of IT. And to demonstrate the concept, I will use "transport" as a metaphor (and thanks to J.P. Garbani for introducing me to this metaphor!).

So you are the "transport" department in a company, and the business comes and says they need to get from A to B. Your options are:

  • Buy them a car
  • Lease them a car
  • Rent them a car
  • Buy them a bus ticket

In the IT world, someone comes and asks to get from A to B. What do we do?

  • Put in and integrate a new application – this is the equivalent of BUILDING a car from scratch. We buy the components, the engine, the wiring, the wheels, etc., and put it together for them. What they get is something that vaguely resembles a car
  • Add to/modify an existing application. This is the equivalent of putting a new engine/tyres/speedo, etc. on our car-looking thing
  • We then might outsource this solution to someone else to manage for us – which is the equivalent of leasing the car
  • We might buy a hosted service – which is like renting a car
  • Or we might go to the cloud – which is the transport equivalent of buying them a bus ticket

Either way, we can't do the IT equivalent of BUYING THEM A CAR! Why not? Well for years everyone has said that people's requirements are different, so everything needs to be customized. To that argument, I say look at the car industry. There are over 20 different manufacturers/brands out there. Within each brand there are probably 5-10 models. Within each model there are 3-6 variants, within each variant there are 15 colors, 5 trims, 5-50 accessories, etc. The auto industry has mastered the art of mass customization. Why can't we?

I believe that within a few years, we will be able to buy "cars" in IT. These will be fully integrated, all-encompassing solutions that will be able to be mass customized. Instead of taking 6-12 months to set up a call center, we will be able to buy a Call Center Machine/Appliance. Instead of taking 2-3 years to implement an ERP or finance solution, we will be able to buy an Order To Cash Machine. These machines will include the hardware, the applications (pre-integrated), the middleware, operating system, management software, etc. They will be complete solutions that have some pre-customization available, and they will just work. And like cars – when they break down we take our Mazda 6 back to Mazda to get it serviced – with these solutions there will be one point of service and management. The one vendor will manage everything. 

And I can hear (and have already heard) the arguments. Why would you buy this internally if you can source the exact same solution from the cloud? This is a good point – and some people will source from the cloud. Others, however, will have concerns around security, compliance & regulation, etc. Mid-sized firms, as a general rule, like to see/touch/hug the solutions that they spend money on. So these companies may be more attracted to this model. This model of sourcing also is attractive in those countries that allow organizations to write down capital purchases against their tax liabilities – something you cannot do with cloud as it is an operational expense. So for some companies, this model will make more sense financially than the cloud model.

The implications of this model are huge. What happens to all the point product vendors (i.e. CA Technologies, BMC etc) whose solutions are no longer required? Could this model drive further consolidation between the IT vendors? How will this impact outsourcing arrangements?

The next question is "which vendor/s are positioned in this space"? Well there is one company with the hardware, operating system, database, middleware, applications, and management software, and that, obviously, is Oracle. They could definitely drive adoption of this model – and get a huge competitive advantage against their major competitors (Microsoft, SAP, and IBM). You could see them making the product hugely competitive on price compared with buying the separate components from IBM, HP, CA, Microsoft and SAP, as they could give the hardware away for free, with the knowledge that the 22% a year maintenance fee will recoup that in the second year. The other important point to note is that they do not rely on consulting & integration services for a large chunk of their revenue. Will they do it? I believe they already have, to an extent, with their Sun Oracle Database Machine. I know it does not encompass an entire business process, but it is a good first step to the world of "business technology machines."

At this stage my thoughts on this new model of IT is a work in process. I would now like you to pick holes in it, tell me where I am wrong, tell me where I am right, and tell me what I have missed. The more refined this model becomes, the easier it is to pick the winners and losers, and the better advice we can give you in your IT to BT sourcing decisions. Feel free to comment below, or e-mail me at, or send me a tweet @timbo2002. I will be building this model out over the next few months, so the more feedback I can get, the better. If you will be at the Forrester IT Forum (or the CIO FLB meeting) in Las Vegas next week (26-28th May) and would like to discuss in more detail please drop me an e-mail and we can meet up there.