Cloud Musings

Can the cloud computing model transform service delivery?

The major software vendors are pushing at pace into cloud computing. There is inevitably much talk of data centres and ‘what if’ scenarios, but most ‘should I put my company data in the cloud?’ questions have been, or will soon be, resolved in our minds as we shift towards a new computing model.

What I want to explore here is the bigger opportunity for using the cloud model to change the way we deliver business systems.

Take up the slack

I once read a case study somewhere in a management training guide for eternal navel-gazers. It was about a company that received an order to deliver double its normal production output for, I think, pallets. When objections had subsided, the company went ahead; it took the order and attempted to double its production output for a week.

At the end of the week, it had constructed and delivered twice as many pallets. Success. But was this down to inspired management, a brilliant workforce or simply enough slack that could be taken up to meet the demand?

The case study went on to explore all three possibilities. I’ve forgotten which publication it was and can’t remember which conclusions were drawn. But I’m willing to bet that a good deal of the company’s success was down to the simple fact that it had built in slack or latency.

Skilful business managers optimise latency. The very skilled lend it out to others when they themselves don’t need it. Others simply have latency because that’s how it worked out.

So, what does all of this talk of latency have to do with cloud computing?

The cloud gives us a chance to control closely how much spare computing capacity we have. If we are proactive, it allows us to drive it right down to the point that we are no longer carrying any spare capacity at all.

Great news. And Microsoft executives should be pleased with their efforts to make this possible. From my perspective, I can now offer our customers some fantastic solutions. I will never again have to pretend to be interested in redundant arrays or inexpensive disks and vast quantities of computer hardware, only ever used in huge upswings in volume or catastrophe. Latency, waste.

So, I think its all good stuff. We buy our computing as we need it, and it’s a pretty sure bet that it’s a whole lot cheaper than we could do it for ourselves. From experience, it has also been more reliable. But how do we continue to innovate and push our companies forwards?

A new model for service delivery

How will we provide the people skills around cloud computing? More importantly, how can we use the paradigm that cloud computing gives us to reduce latency to a minimum in the services we use to build new systems? Can we innovate and take our businesses forward?

Can we put people in the cloud?

I’m not advocating it, but it might be worth a try (many IT folk I know might be better suited to cloud living!).

I’ve worked for large outsourcing companies and am predictably cynical about the perceived benefits: At the most basic level, when organisations outsource their IT functions, people and equipment are organisationally shifted from one place to another but, in reality, not much changes.

The outsourcer can usually win a bit of additional business by charging for ‘extras’ and, in my experience at least, then puts its arms tightly around the IT function and strangles it – turning it from a centre of business enablement to a centre of business cost.

As a small company that has a good deal of business in small outsourcing contracts, I don’t condemn them completely: they can provide real value as long as some of the effort focuses on latency carried by the organisation

Cloud computing does this for software and systems. But how can we adopt this model for service delivery? We’ve shifted the tin, software licences and the operation to a world populated by cumulus, cirrostratus, nimbostratus and the like. In theory, we’ve reduced our latency in this area to zero. What is the latency in system delivery?

Can we provide IT services on a subscription basis?

Look at your typical medium sized business: it will probably have an IT department with developers, project managers, networks people and a myriad management people. How much latency is there in that delivery function? 5% sounds okay. What about 20%? Maybe 50%? Possibly a little on the gluttonous side of reasonable.

I’m not talking about everyone creating more value all of the time: I’m suggesting that these functions have, and always will have, significant amounts of built-in slack. While the organisational design stays the same, and IT remains so granular, it’s hard to see how this can improve.

We need to reduce the system development footprint. We need to increase the bang for your buck ratio. How long will it be before our business masters say, ‘Well, Microsoft has brought down the cost base of my business computing but you guys and girls are still charging me the same to deliver applications’?

It’s a fair comment that we have worked with our clients to address. Over the coming months, we will be discussing the services model for subscription (cloud) based computing and we’d welcome your thoughts.

About the author: Ben Gower is CEO of Perspicuity and has been working closely with Microsoft to formulate its cloud computing strategy. Perspicuity is one of a small number of specially selected Cloud Accelerate partners. Office365 is a packaged offering of Microsoft Online Services which encompasses SharePoint, Exchange and Lync