Last month, I've read a post by James Governor of RedMonk (an analyst firm specialized enough to actually know what they are talking about :) where James talked about the “best practices” vs “best infrastructure” inquiring whether SMBs would be better off using infrastructure provided by large service providers like Google, Yahoo, Amazon, etc. and if SMBs do choose to use these services more and more what would be the impact of this in the industry.
In the comments, I had pointed out the differences between the “best infrastructure” and “best service (support)”. I keep running into this theme so I wanted to write more about it, if for no other reason to clarify my own thinking.
Providing "best infrastructure/service" is expensive and difficult. It requires significant resources and organizational maturity that is elusive. Very few organizations manage to overcome the organizational challenges that impede service oriented management (ITSM, BSM, etc) and change the way they work. (read Doug McClure's post that I mentioned in my previous post).
The “best in class infrastructure providers” of today (read Yahoo, Google, Amazon, FedEx, etc) do have the best infrastructure. They have the size and the resources to optimize everything. Simple economies of scale (in the extremes). They have built very scalable infrastructures. They do not incur drastic cost increases for providing services for significantly more customers.
There is no doubt in my mind that there are significant advantages for (especially) SMBs to shed some of their internal IT services and use services from best in class providers. But currently there are obstacles preventing mass adoption of these services. The “best in class infrastructure providers” do not provide the best support, at least not yet.
If you work with one of these giant service providers, life is great when everything is working, but when something goes wrong (and it always does), your options are limited. There is typically no SLA, and support is minimal. It is hard to get someone to talk to and problem resolution times are unacceptable for most businesses.
This is not all that surprising. Traditional ways of providing support is still very resource intensive and does not scale well. Building scalable support structure turns out to be just as hard if not harder. Innovation in this area is desperately needed, and I think whoever solves the support problem will lead in this field. I believe there is also an opportunity for startups in this area.
How do we enable customers to monitor, measure and manage the services provided by external service providers? Where does this fall in Business Services Management? Are current approaches to management, strong emphasis on massive implementation of CMDBs, (as it seems to be what' promoted by major vendors) suitable if businesses increasingly use infrastructures provided by external providers?
