Enterprise software is not a commodity

This post was prompted by a recent blog I read – Refuse to be a Cloud data hostage. The point of this article is that businesses should treat their cloud software like a commodity and be able to switch as they please. This stems from the argument by the Chief Technology Officer of Amazon, Werner Vogels, that:

“You should keep your providers on their toes every day. If we are not delivering the right quality of services, you should be able to walk away. You, the consumer of these services, should be in full control. That is core to our philosophy.”

When I read this, the word that stood out was “consumer”, not business. Consumers may have the privilege of being able to jump whenever they want, but my view is that businesses do not, and perhaps should not.

If you’ve ever worked in a manufacturing organisation, you will know that they invest in machinery only after a long evaluation process. Why? Because it is expensive, they may have to make changes to their processes in order to benefit from it, and they have to train staff to use. Then, if they want to keep getting benefit, they partner with the machinery manufacturer to maintain and improve it.

That is what should happen with enterprise software, whether it is in the cloud or not. Enterprise software is not a commodity, it is likely expensive machinery, and should be treated as such. If you think enterprise software is a commodity, then you are going to get very little benefit and far less than you had expected.

where is Dynamics in the cloud?

Jason Hiner at Tech Republic published the PowerPoint slide from TechEd 2010 that showed Microsoft’s strategy for private and public clouds. I think there is something missing from that picture – on the private side you can see Dynamics, but it is not on he public side. So my questions are: where is Microsoft Dynamics, and especially the ERP component, in Microsoft’s cloud strategy? What does this say about Microsoft’s plan for ERP Software-as-a-Service?

New platform for the new decade

I read with much interest Don Dodge’s predictions for 2010 and the new decade. I find predictions a bit of a waste of time – when you look are what people predicted for 2009, about half came true, which is what a random selection would give. However I was struck about the prediction on future computing and mobile computing.

“Your cell phone will become your primary computer, communicator, camera, and entertainment device, all in one … I think in the near future there will be docking stations everywhere with a screen and a keyboard. You simply pull out your phone, plug it into the docking station, and instantly all your applications and data are available to you … Your phone will have enough storage so you can decide which applications and data are stored on your phone, and which will be in the cloud.”

When I look at what I can already store on my cellphone, this prediction seems quite plausible. But if people store applications on their phone, which they will presumably choose themselves, and if they decide to use the cloud to select their preferred applications, how will this impact the role and responsibilities of the company IT function? Their role of deciding what applications are suitable for the organisation becomes irrelevant, but they still have to ensure application and data security and integrity.

 ”Mobile phones are clearly the next computing platform … Mary Meeker of Morgan Stanley says Mobile Internet usage is bigger than most people think, and it is exploding.”

Application developers will have to re-consider the presentation layer for a different user interface and experience, and will have to assume that the mobile interface will be the preferred or default one, rather than as a side issue.

Where I do have a problem is the predictions about cloud computing and mobile bandwidth:

The explosion of reliable broadband bandwidth, virtualization technology, cheap storage, memory, and servers, has made Cloud Computing the obvious choice for the next decade … Why buy servers, hire IT admin to manage them, buy operating system licenses, application licenses, pay 20% maintenance fees every year, worry about security updates/breaches, hassle with asset management, etc., when you can just “pay as you go” with cloud computing resources? …
The new 700Mhz wireless spectrum became available in 2009, and will be built out over the next decade … Cell phones will see the same explosion in bandwidth in the coming decade, which will enable new applications and uses.”

This is a simplified view of the world. Firstly, business applications for even small and medium businesses are getting more complex. I’m not talking about simple accounting or CRM solutions, but the complex applications to manage orders, receive and dispatch inventory, schedule and manage manufacturing operations. It isn’t as simple as paying and starting, as the “pay as you go” mantra likes to make out. It may well be that business hands over the management of the application infrastructure to a cloud provider, but every successful business has a particular way of working that is different to others, and that is unlikely to be assisted by a standard enterprise application.

Secondly, the comment about mobile bandwidth is US-centric. Developing countries like South Africa are constrained in all sorts of bandwidth (Internet, radio, cellphones) because they have been already allocated to developed countries years ago. That is not to say that cellphone bandwidith will not significantly increase. What Don fails to note is that the growth of the mobile Internet will probably come more from developing regions like Africa than the US, as reported by Opera and Google.

However you look at it, the next decade is probably going to redefine the way we use, consume and interact with computing resources.

Enterprise cloud computing – mea culpa

It looks like I was mistaken in my previous comments about enterprise cloud computing. I saw a blog by Irving Wladawsky-Berger from March this year in which he points out that BusinessWeek had an article about cloud computing, and that IBM had started a project in 2007. So it’s not such an innovator phase technology as I thought it was.

But I maintain that if I asked a sample of business people in the small- to medium-business space, most of them would not have heard of the technology; it’s still very much in the early adopter phase.

Enterprise cloud computing in the adoption lifecycle

I have been in IT in a developing country long enough that I am now initially sceptical about new technologies that get touted from the US; that country has a belief in technology that sometimes amazes me. One of the latest technologies is “cloud computing”, and Oracle and Intel have announced a project to develop the enterprise readiness of cloud computing saying that they want to provide business with “flexibility of choosing to run their enterprise systems in either private or public clouds.”

Without a generally accepted definition of cloud computing, I am prepared to use Oracle’s definition as ”the notion of providing easily accessible compute and storage resources on a pay-as-you-go, on-demand basis, from a virtually infinite infrastructure managed by someone else.”

While Oracle and Intel have started on the research, I reckon 99 percent of the population wouldn’t have a clue what cloud computing is, particularly the business community. Therefore, cloud computing in the technology adoption lifecycle is definitely at the innovators phase.

According to a recent Sandhill blog, cloud computing is at the top of the Gartner hype cycle, which means to me that the problems, failure stories and disbelief have yet to come out. According to Gartner, for vendors and business customers ”the key activity will be to determine which cloud services will be viable, and when.” Sandhill makes some pretty obvious recommendations for vendors:

Clearly spell out the cloud’s value proposition for customers

Eliminate friction between cloud vendors and components

Make cloud services easy to use, purchase and deploy widely