Another Take At Integrated Vs. Best-of-Breed

I was locked away in meetings most of this week, and didn't have a chance to participate in a flurry of posts and discussions on this important topic.

Scott Lowe got the latest wave going with his prognostications on a certain vendor, Marc Farley added a thoughtful post, and Chris Mellor joined in the fun, as well as Storagebod.

But the topic isn't done yet -- we'll be arguing this one for decades to come.

Working as I do for EMC, I'm free to sign up on either side of the debate: EMC offers best-of-breed products for consideration (insert long list here), or conversely EMC is part of an integrated stack (e.g. VCE and Vblocks for private cloud deployments).

So, I'm going to wade here with a decisive statement: both perspectives are right, but what changes over time is where the boundaries of the debate occur.

In A Nutshell

If I could paraphrase the positions, they go something like this.

If you're in the best-of-breed camp, creating value in IT is all about differentiation.  You can't get differentiation without best of breed, ergo, if you are to be a differentiated IT organization, you must adapt a best-of-breed strategy.

If you're in the integrated stack camp, creating value in IT is all about delivered the best services for the lowest total cost.  IT can't usually add much value by re-inventing the wheel.  Focus your differentiation efforts where they create value.

Like most interesting debates, there's a germ of truth in both positions.

What I'd like to do is add another dimension -- how does this change over time?

A Personal Example

I've been using home PC tech for a long time.

My first wave of home-build involved a ton of research: best processor, best memory controller, best bus implementation, best networking card, best graphics card, best cooler, best sound card, best power supply, best case, best storage and so on.

Pick the best parts, put them together, spend a *lot* of time making them work together (if ever!) and live with the results.  Lots of fun doing the research and assembly; a royal PITA when things didn't work as they should.

My second wave was a bit better: pick the best motherboard, a few storage devices, maybe a graphics card, and I'm good to go.  A lot less time doing research, less integration work, and a more usable and stable system.  Less geek fun, but also less hassle.

My third wave basically involved walking into a BestBuy, picking something out of stock, and maybe upgrading it with a gaming card.  Almost no integration or support time, and a very stable system indeed.  Time-to-value: about 15 minutes.

And my nascent fourth wave involves buying an iPad, handing it to a family member, and calling it a day.  More value, less work.

I don't think I'll be buying another traditional desktop or notebook for a very long time.

Technology Evolves

Hopefully, my short story here captures a key point: as the technology evolves, differentiated value moves up the stack.  Less value comes from the underlying technology, more value comes from how it works together.

Does value still come from differentiation?  Yes, but the opportunities to do so inevitably move "up the stack" for most situations.  And it appears to be an inexorable trend, like commoditization or open source.

Using storage as an example, EMC's storage business evolved from (a) supplying and supporting only the array, to (b) supplying and supporting the storage network, to (c) supplying and supporting the HBAs, drivers and path management for the storage network, to (d) providing software to manage, control and optimize storage in different ways, to (e) even offering storage as a managed service rather than a technology stack.  The degree of integration increased, the value moved up the stack.

This discussion isn't entirely limited to infrastructure, either.  I was around when the big SAP ERP wave hit in the mid 1990s -- from standalone business apps (remember Accounts Payable) to integrated business software.  The value had moved up the stack.

Do most IT organization hand-build their own servers and storage from components and software?  A few do (think Google) because they believe they can create unique value by doing so.  But the vast majority of us buy our technology in pre-fab, pre-integrated chunks.

As an example, when I discuss something like a Vblock, I position it as an alternative to the traditional do-it-yourself model.  Sure, you could do all the work yourself to select the best components, and do all the integration and support yourself, but why?  I know, you enjoy building your own PCs ...

At the end of the day, don't you just want a pool of virtual resources to throw at whatever comes your way?  Why not just buy that, rather than have to build it yourself?

Evolution Causes Disruption

When things change, people are affected. 

If you're a specialized technology vendor (pick your favorite discipline), you don't like this particular secular trend, because it doesn't work in your favor.  Hopefully, you can claw out enough market share to be attractive enough to be acquired, or perhaps invest like crazy in partnering with everyone else's integrated stack. 

Make no mistake, for smaller vendors in markets that are rapidly commoditizing and being consumed into integrated stacks, this is an argument that they'd prefer to win at all costs.

If you're on the other side of the fence -- a specialized technology professional (pick your favorite discipline), you may resent the fact that your expertise is being integrated with (and consumed by) other disciplines, leaving you less turf over time that you can claim as your own.

Hopefully, you can evolve and integrate your personal value proposition with others around you, otherwise you will find the demand for your services declining over time as the world changes around you.

Putting It All Together?

Does value creation come from differentiation?  You bet -- no argument there.

The real discussion for all of us going forward is -- how will the places for differentiation change in the future?  What will become integrated and/or commodotized, and where can we add value -- not only for our organizations, but ourselves personally -- as the landscape changes?

Interesting times indeed ...