So many people enamored with the brave new world of web-scale IT. You see them online, at conferences, etc. — brimming with passion, enthusiasm and excitement. I call them the clouderati — their heads are clearly in the clouds.
It seems that every few weeks this group brings us a new shiny meme that captures a lot of attention: containers, devops, OpenStack, NoSQL databases, Open Compute — even machine-learning algorithms for energy efficiency.
I’m certainly not immune to the attraction. So many innovative concepts to learn about, evaluate, debate, etc. If you’re looking for intellectual stimulation, there’s no shortage.
But, in many ways, it’s world unto itself — this world of the clouderati.
A shiny bubble, floating over the vast and often grim landscape of real-world enterprise IT. No one is debating the coolness of whatever the latest idea might be; it’s just that it’s not consumable by the vast majority of enterprise IT shops.
If it’s not consumable, it’s not particularly interesting to this audience.
Judging by many of the conversations I’ve had, I believe that many of these enthusiastic new web-scale tech people might not have a gritty appreciation of the enterprise IT terrain. So, in an effort to promote a more productive dialog, here is my quick primer on just a few of the harsh realities of enterprise IT.
#1 -- Business Model Matters
If you’re a web-scale company, IT doesn’t just support the business, IT is the business.
Most web-scale outfits have business models that prioritize making IT differentiated, so they invest like crazy in getting better at it. They have large teams, they hire PhDs, etc. As a result, they can experiment, they can write code, etc. — all part of their model. As it should be.
By comparison, most enterprise IT organizations work for businesses that don’t prioritize IT differentiation. So many executives see IT as little more than a cost of doing business, rather than a source of potential business innovation. Thankfully, that’s starting to change.
When comparing web-scale IT and enterprise IT, the application profiles are very different. Most enterprise IT organizations have to deal with wildly diverse application workloads — each with unique requirements vs. the relative conformity of most web-scale environments.
Yes, Google does it. But enterprise IT is not Google.
Fun exercise: if you have the opportunity, ask a web-scale company where they run their boring-but-critical applications like financials, HR, CRM, etc. Hint: it usually isn’t on their cloud.
#2 — Perspective Matters
The vast majority of people I know who work in enterprise IT have a difficult and often thankless job. They work long hours. They rarely get to go to conferences and seminars unless they’re local events.
They live in a rough world of office politics, budget constraints, dodgy environments, conflicting priorities and co-workers who — politely put — may not be bringing an “A” game to the table. Job security and career advancement is frequently optional.
I’ve found that the vast majority of enterprise IT people don’t pay attention to most of what’s being talked about in the cloud world — simply because it doesn’t relate to them and their world.
#3 — Legacy Matters
In most enterprise IT environments, they’ve already made huge investments in certain technologies: infrastructure, software, applications, etc. Not only have they spent a boatload of their scarce IT dollars on this tech, they’ve invested considerably in training their people how to use it effectively.
I encourage people to present their new tech in the context of how it works with what people already have, instead of assuming a convenient fresh start. While it’s true that a few fortunate IT groups have the ability to stand up legacy-free test beds to try out new things, that tends to be the exception rather than the rule.
#4 — Skills Matter
IT organizations are basically collections of people, each with a certain skill set. Indeed, instead of looking at the technologies a group has invested in, consider the portfolio of skills needed to make it all work.
Certain propositions that demand a critical set of very new and hard-to-obtain skills (e.g. devops, big data analytics, etc.) are simply non-starters for so many IT organizations, as they present a skills hurdle that seems to be near insurmountable.
Look at it from an IT leader’s perspective.
They’ve got only so many headcount. They’re not going to get any more anytime soon.
If you ask them to invest big in a new skill set, that means de-investing in something else that’s needed to keep the lights on.
And that’s a big ask.
#5 — Supportabilty Matters
When something breaks (and it always does), it’s an incredible fire drill between upset users, limited IT staff and the supplying vendors. A well-established and well-supported technology will always be strongly preferred to one that isn’t — even if the new shiny thing is better, faster, cooler, sexier, etc.
That’s just one of the reasons larger IT vendors have such a clear advantage when it comes to enterprise IT — they understand just how important this is to customer.
As an example, being “open” is often less important than having the ability to support your enterprise customers when the going gets rough.
#6 — Relevance Matters
“If it ain’t broke, don’t fix it” — words to live by in an enterprise IT setting. I find that so many of the newer propositions just aren’t all that relevant to what enterprise IT people consider important.
The first category is “your users are demanding it”. Put mobile, big data, etc. in this category. Sometimes that new requirement comes with budget, sometimes not.
The second category is “we have to continually improve at what we do” — speed, costs, quality, etc. — and usually precious few resources available to invest in continual improvement.
Unless whatever is being proposed clearly falls into one of these two categories — and doesn’t come with a bunch of external complications and costs, it’s probably a non-starter in the enterprise IT world.
Enterprise IT Isn’t Going Away Anytime Soon
As John Maynard Keynes famously pointed out, “in the long run, we are all dead”.
While we’re waiting for utopia to materialize (built of unobtanium, naturally), we are left with an enormous opportunity to help enterprise IT organizations get better at what they do.
Building Bridges, Not Walls
So many of the new web-scale tech vendors seem to be putting up walls to adoption by enterprise IT organizations. Maybe that’s intentional (e.g. knowing your target market), maybe that’s inevitable. Maybe selling to enterprise IT organizations just isn’t sexy enough for some people. I don’t know.
There are certain IT vendors (my employer VMware included) that think in terms of building bridges between the existing enterprise IT world, and the emergent world of web-scale IT.
The idea is to give people technology that makes their existing world a better place (e.g. virtualization) and then give them a defined path to walk to a new web-scale world — when they’re good and ready to, that is.
... But Please Don’t Stop
As I’ve said initially, all the new tech that’s being created as web-scale emerges as a distinctly differentiated form of IT is very cool, and is an impressive engine of innovation. I’d hate to live in a world where we didn’t have the benefit of all those cool ideas to consider and understand.
Just don’t expect them to be adopted in most enterprise IT shops anytime soon :)
Like this post? Why not subscribe via email?