Here is a post from 3PAR CEO David Scott suggesting that my ideas in the Storage Killing Fields post were incorrect:-
————————————————
As a self-declared [:)] “best of breed” primary storage manufacturer, let me try to put an alternative position forward relative to your recent article on StorageWhimsy:
articlehttp://storagewhimsy.wordpress.com/2010/06/18/the-storage-array-killing-fields/
I decided not to post it because it was too long, but feel free to use it any way you wish if you feel it would usefully add to the debate. [I think posting it 'as is' is the best way to represent David's view.]
It all rests on how pervasive one views the future of the delivery of enterprise IT as a utility service in the form of public clouds (IT-as-a-Service — IaaS/SaaS). I’m going to use your analogy, but it would work better if we were used an analogy about the mobile phone business which is a true utility service model (think AT&T and Verizon in the US).
Suppose you view the car manufacturers (e.g. BMW, Mercedes) as the new public cloud IT-as-a-Service suppliers (a car after all is a standardized service, since the consumer doesn’t get a lot of choices as you point out). Their business is to create a best-in-class infrastructure (the car) that attains the particular service level (performance, comfort, features) at the most competitive price point they can achieve, and then lease (utility pricing model) or sell the car.
BMW is in intense competition with Mercedes and they are in an evolutionary arms race to out-compete each other from generation to generation on these terms. They integrate car seats, steering wheels, electronics, pedals and everything else in the car from a set of product manufacturers to meet this aim. Neither BMW, nor Mercedes, can afford to choose anything that is less than Best-of-Breed (meaning whatever capability/cost combination is “optimum” for the type of car and target segment they are trying to attain) otherwise they will lose in competition to the other. Ultimately, if they continue to handicap themselves by choosing sub-optimized components they may lose badly enough that they will be acquired.
The systems vendors in this view of the world would be the equivalent of a car component suppliers trying to roll up an “integrated stack” of car seats, steering wheels, electronics, pedals etc. and marketing them to BMW and Mercedes. Some of the components are Best-of-Breed, but many aren’t. However, each of these systems vendors tries to claim there is some integration benefit from buying everything from them. It’s a superficially attractive argument.
Now let’s say BMW buys such an “integrated stack” from one of these car component roll-up vendors called Fully Unified Distribution corporation (FUD). BMW would inevitably put themselves at a disadvantage to Mercedes because some or many of the components they are forced to integrate from would be sub-optimized (FUD has great car seats, but not the latest electronic traction technology, nor iPOD playing systems – just CD players). Mercedes, continuing to choose Best of Breed, would not suffer the same handicap, and would with almost all certainty start winning in the car-buying consumer market place. Mercedes is able to integrate better functionality components (great car seats, latest iPOD playing systems, and optimum electronic traction technology) while BMW had to accept whatever FUD had in their “integrated” arsenal. And Mercedes would even find themselves being able to lease their higher functionality cars, at a cheaper price than BMW, because FUD was exerting undue price control on BMW’s procurement department due to BMW’s dependency on FUD’s integrated stack (proprietary lock-in). BMW loses. Mercedes wins. BMW’s never going to let that happen, just as no IT-as-a-Service provider is going to let it happen. If you view the enterprise IT world as moving increasingly towards IT-as-a-Service then a Best-of-Breed strategy will be seen as the key to success in the public cloud. (BTW, if you counter by saying that BMW could choose, only the Best-of-Breed components that FUD has to offer and then insert other Best-of-Breed components from alternative suppliers, you have effectively said that BMW has maintained a Best-of-Breed strategy like Mercedes.)
Even traditional enterprise IT organizations who can successfully build their own private clouds because they have appropriate IT competency and scale, will find themselves having to compete with IT-as-a-Service suppliers for their own internal business units’ projects. There is a clear trend for internal business units to increasingly ask for bids from both ITaaS suppliers as well as internal IT when considering new projects. This new type of competition will force the traditional enterprise IT organizations towards Best-of-Breed strategies as well s they turn themselves into Internal Service Bureaus.
And if an enterprise IT organization doesn’t have the scale or IT competency to build its own private cloud, then it should probably be an organization that moves earliest to become a customer of the IT-as-a-Service providers themselves (implemented on Best of Breed stacks). After all, if they have never seen IT as a core competency, why buy an integrated solution when there is a newer, better, renting alternative?
Where’s the room for integrated stacks in this picture?
What integrated stacks really represent is a defensive competitive strategy to slow the transition from traditional enterprise IT to the IT-as-a-Service world by getting enterprises that don’t have strong IT competency or scale, to build their own private clouds based on these integrated stacks. The systems vendors have incumbency in the traditional IT data centers; they don’t in the IT-as-a-Service world. The tectonic shift towards IT-as-a-Service represents an existential threat to the systems vendors, so they have to slow it down.
The survival of Best-of-Breed really depends on whether one thinks there is an inexorable shift towards IT-as-a-Service. Are the car manufacturers the “systems vendors” or the “IT-as-a-Service” suppliers?
——————————————————
This is a persuasive and very well-presented view. Thank you David.
