I’ve been inundated with inquiries about cloud computing, from vendors and users. A lot of the discussion on cloud computing is about technologies in the cloud, maturity of cloud services, which workloads/services could be sourced in the cloud. I think there is a major challenge that isn’t discussed enough – how are we going to manage cloudsourcing?
If this was a “simple” outsourcing contract, it’s straightforward: we create a contract, with service level agreements, price, etc., etc. But unlike traditional outsourcing, cloudsourcing will be extremely dynamic, and extremely granular (unless you believe that everything will filter through a handful of megaproviders, which I don’t). Sourcing choices won’t be made once every few years, they could be made daily. Even worse, cloudsourcing will not just require finding a service in the cloud that meets business needs for a particular situation, it will likely be about logically doing systems integration – services integration – between many potential cloud service providers.
And the dynamic nature of cloudsourcing won’t only be reflected in internal decisions to use the cloud – the cloud itself will be dynamically changing. Today, one provider might have the best price, the appropriate quality of service. Tomorrow, their costs might rise, they might incur an outage, they might go out of business. Somehow we need to manage these dynamic decisions and the dynamic turmoil in the cloud to our best advantage.
Another issue is that business customers will be confronted with the same kind of opportunity they faced in the early days of client/server – IT says “no”, so the business customers went out and bought their own servers and packaged software. If business customers start making these decisions in the cloud, since they often don’t understand their service level requirements to the level of detail needed to cloudsource, the failures will be rampant.
I believe there are two entities that will be created to respond to this: in large enterprises, they will create a dynamic sourcing team that makes day-to-day decisions about sourcing. They orchestrate the services in the cloud to meet business needs. This team needs to be business- and technology-savvy – a rare mix of skills that we need to generate to cloudsource effectively.
Smaller businesses are even more likely to rely on cloudsourcing, but they will not have the skills needed to orchestrate cloud services efficiently. For them, a breed of service brokers will emerge. These service brokers will be an evolution of today’s VARs/resellers/systems integrators. They will take responsibility for the overall service level requirements in the business. They will likely be skilled in specific industries (and perhaps their unique regulatory requirements, etc.). They will be able to monetize their value by having deeper skill in the cloud market than small businesses can muster, and by leveraging a rapidly changing market to continue to find the best deals to keep costs as low as possible (and still meet service level requirements).
On the opposite side of the equation are businesses who are thriving in the cloud and the connections enabled by the cloud. For them, being “discovered”, finding a market need (even temporary) and filling it – this is how they succeed. It will be the challenge of dynamic sourcing teams and service brokers to sift through the noise and find the best providers – and orchestrate their services.
Category: Cloud Tags: cloud computing

Thomas J. Bittman





































































































6 responses so far ↓
1 Cloud Computing for Early Stage IT « Early Stage IT Blog December 18, 2008 at 12:00 pm
[...] to negotiate and manage contracts. According to Gartner analyst Thomas Bittman, one of the major unrecognized challenges of cloud computing will be how IT departments will manage the [...]
2 Deep Bhattacharjee December 29, 2008 at 6:37 pm
Do you think it will be easy to lug terabytes of data around from one cloud provider to another? Lets say a company now has its data in Amazon’s simpleDB format. How are you going to move that data to some other cloud for computation without converting the data to a format that the other cloud provider understands. What you say may happen if the protocols and data formats are standardized across various cloud providers. That will not happen as the early leaders would like to keep their market share. I am not sure how long it took before all electric utility companies supplied power at 60 Hz. Besides the electric grid is a bit different from the internet.
3 Tom Bittman December 31, 2008 at 1:52 pm
Deep – agree, but the analogy is closer to the server market than the electricity production and distribution market. Early servers – very proprietary, very hard to “leave” – but as competition moved from lock-in to richness of offerings, new workloads tended to be put elsewhere. In the cloud, standards and openness (as always) will be driven by the little guys – the medium-sized service providers who work together to compete with the big guys. It’s never a perfectly competitive world, but the tendency will be in that direction, and I believe the speed of the shift will be lightning-fast compared to the electricity production market, and much faster than the server market. Near-term, the question won’t be moving terabytes of data, it will be choosing to source a new and separate app with a provider that enables choice/competition/exit. Let’s talk in five years!
4 How The Cloud is Changing IT Services « Control Group Blog July 26, 2009 at 12:24 pm
[...] event with Google and Mozy that we have dubbed “CloudSourcing”, taking a note from Gartner and tweaking it a [...]
5 Louis December 13, 2009 at 10:51 am
What would be the skill sets needed for a dynamic sourcing team and service brokers?
6 HP Blogs – Cloud Computing and Lateral Thinking – The HP Blog Hub | PLI Cloud September 24, 2011 at 2:32 am
[...] Managing Cloudsourcing. [...]
Leave a Comment