Many of us have watched trends wash over the industry landscape. Some come and go quickly; others make more substantive changes.
I've had the unfortunate privilege of observing many different waves hit, and I'd like to think I can learn a bit from each one.
As we sit here in 2013, it was only four short years ago that the first vituperative cloud arguments started to erupt online: true clouds vs. fake clouds, clouderati, accusations of cloudwashing -- remember all that?
From my perspective, the root cause was simple: collectively we were using a single term with multiple meanings -- and thus being perceived very differently depending on your viewpoint.
History may not repeat itself, but it does rhyme: the same phenomenon was evident when the big data conversation erupted a few years back. Once again, a single term was being stretched and tortured to mean very different things to very different audiences. Frustration and cynicism is an inevitable by-product.
I suppose the most recent phenomenon is software-defined anything.
Those of us who are vendors and industry-watchers historically have matters much worse for everyone else. We focus on one shiny aspect or another of the discussion, and often don't take the time to construct a broad enough perspective for the less-entrenched participants.
Since IT is an industry powered by big ideas, I think we're doing ourselves a disservice. People can't act on a concept until they understand it -- in their terms, not yours.
If you followed the cloud debates early on, the discussion broke into three very distinct camps.
The second camp saw cloud as an operational model -- run your IT delivery function using these kinds of processes and automation, and -- presto! -- you have a cloud.
And a third very vocal contingent insisted it was all about the consumption model -- if it came over a wire and was variably priced, then de-facto you were using a cloud.
Who ended up being right? All of them -- and none of them.
You see, I found quickly that each and every group I talked to -- enterprise IT, service providers, business consumers of IT, etc. -- was looking at the problem through their historical lens, rather than through all three aspects.
As an example, the enterprise IT groups would think "well, I'm doing virtualization, that's cloud, isn't it?". Service providers would point to their processes, and do something similar. And, of course, business users would focus solely on the convenient consumption model.
You had to start the conversation with an area they were familiar with, and then bring in the other aspects. No other approach worked well.
Not even a NIST definition :)
As I sit here today, there's general consensus that cloud involves three different aspects: technology, operations and consumption. Comparatively, the technology piece is the most straightforward: lots of decent options and packaging depending on your tastes. Changing the IT operational model -- well, there's people involved -- so that requires a bit more effort. But it's well-understood these days.
However, teaching the business to become a proficient consumer of on-demand IT -- well, that's very heavy lifting :)
My Second Example -- Big Data
It must be my sampling bias, but when the big data discussion first started, it once again seemed so very technology-centric: Hadoop, R, scale-out storage, powerful analytical tools -- all of that. Quickly, though, the discussion morphed into the need to create new operational workflows and capabilities -- also involving a new class of skills.
It wasn't long before the elephant sitting in the room appeared: we as organizations needed to learn how to use these new capabilities to change the way we made decisions and did business.
I'd offer up the same general state-of-the-union as to where might be today. Big data technology (in all its aspects) is readily at hand. Sure, there's plenty of room for innovation and improvement, but the necessary tools are available and not hard to procure.
Creating the operational processes (and inserting the right skills) can be challenging, but there are plenty of resources available to help you do that, or various packaged services available for more narrowly-focused domains.
However (once again) the real challenge is teaching the business to become proficient consumers -- and act upon! -- the analytical insights that come pouring out as a result. And that (once again) is going to take some time.
Could there be an echo in the room?
Does A Pattern Emerge?
I've got a few other examples, but I'd like to get to some of the key points I feel need to be made.
- In each case, we're really should be talking about a journey of organizational proficiency -- and not a shiny new technology buzzword.
- In each case, it's not just an IT thing, it doesn't create value until the business learns how to use it effectively.
- In each case, we as vendors and industry-watchers tend to make matters more confusing at the outset.
The Self-Inflicted Injury?
As I've said before, IT is a business that's powered by new ideas and new concepts. The simpler and more approachable we make these concepts, the faster things tend to move along. You'd think that we all would be collectively motivated to make things as simple and relevant as possible.
But we -- as a an industry -- we don't do that, do we?
For starters, we go looking for a precise (and hopefully authoritative) definition way too early in the process.
These newer concepts need space to breathe, grow and evolve. Once we attempt to pour definitional concrete on something, we end up debating the definition rather than what it might mean to all involved.
Second, we as vendors (or consultants) are inevitably motivated to stake our claim as to relevancy as early as possible. In addition to producing some occasionally laughable #weaksauce marketing, we just add further noise and confusion to what is already a noisy and confusing discussion.
We also open ourselves up to very justifiable accusations of #cloudwashing or #bigdatawashing or perhaps #SDNwashing.
Third, we collectively spend an enormous effort to re-position the stuff we've always had in the new bright light. Here's a familiar product, now dressed up in cloud. Or big data. Or software-defined something. Let's face it -- these are serious new ideas, and they most likely require new product development (and new services and new delivery models) to be successful.
Hard to imagine that *any* vendor had all the right stuff conveniently sitting there when an important new idea came along.
Fourth, in this era of social marketing, you'll see all sorts of vendors and consultants flooding the airwaves attempting to "join the conversation" with nothing really to say -- other than perhaps parroting someone else's words.
I empathize with anyone who's seriously trying to figure out what's really going on by parsing the social stream. The signal-to-noise ratio can be frustrating.
While these behaviors are understandable in their motivation, they collectively slow the entire process of moving things forward until decision makers can sort things out for themselves.
We make matters worse, not better.
A Trick I've Learned?
I've learned something important from all of this.
Every time one of these new concepts comes along, I pull out an old-fashioned piece of paper and pen, and I draw a matrix diagram of "who cares, why would they care?": customers, partners, vendors, consultants -- everyone who touches an IT discussion from cradle to grave.
I then visit each square, and attempt to establish some empathy with how each individual might see their world. I build up little theories, observations, stories and examples for each.
Once I've decently populated the matrix to my satisfaction, I then go looking for broad themes and ideas that tie the different constituencies together -- hopefully in a common vision that unites instead of divides.
Why? It's that shared vision that can be the most powerful in moving the discussion forward.
At the end of the day, we all want mostly the same thing -- to see people and organizations become even more successful with the powerful technologies that are at hand.
If only we could agree on a few definitions :)