Enterprise IT is a tough job for most practitioners. The list of things that need to be done never stops growing, and resources are always constrained.
Distractions don't help, and there are no shortage of these it seems.
One of the biggest IT distractions I've witnessed -- and perhaps most damaging -- has been what I call "cloud envy".
Maybe call it "Google envy" "Amazon envy" or perhaps "Facebook envy"?
The symptoms are always the same. Folks in IT leadership become enamored with what these web-scale companies are doing with their clouds. They may even visit with them, take the data center tour, all that. They are inevitably dazzled by the experience. They come away thinking, "hey, maybe we could do that too!".
Much time and effort is then lost chasing a dream that isn't right for them: wrong model, wrong motivations, wrong abilities, etc.
Precious resources that could have been spent on things that really move the needle get needlessly frittered away on a fantasy that can't -- and shouldn't -- happen.
We Are Who We Are
I've always been impressed by the time and dedication it takes to be a professional ballet dancer. It's not for everyone. There are very few very good ones, and they focus their whole lives developing proficiency. It's a calling.
Just because I admire ballet dancers doesn't mean that I should try and become one. For starters, I am 56 years old, weigh 220 pounds and a bit uncoordinated.
Yes, I could throw caution to the wind and commit myself to being a professional ballet dancer, but the results wouldn't be pretty, would they?
That's what I think when I meet an enterprise IT group with a bad case of cloud envy - this isn't going to be pretty.
When I encounter these situations, I usually end up sharing a bit of tough love. Not that anyone is going to listen to me, but I feel it needs to be done.
#1 -- They're In A Different Business Than You
All of these web-scale companies share a common attribute: they are in the business of delivering end-user cloud services at truly impressive scale. Their business model dictates their approach to customer-facing IT.
The usual IT vendors didn't build what they need, so they had to invest in building it themselves.
Trust me, if they could have bought what they needed off-the-shelf, they wouldn't have felt the need to invest in truly massive IT teams: application development, architecture, operations, etc.
The scale of these dedicated cloud teams is almost as impressive as the web-scale services themselves. And these teams aren't cheap.
Much like a traditional product company would invest in product design, supply chain, manufacturing, etc. -- these web-scale companies have to either invest in their "factory", or rent one from a cloud provider. Otherwise, they can't get their product to market.
#2 -- Cost-To-Serve Matters, Because Scale
If you have a billion users (or more!) that directly consume your web services, little numbers can add up quickly.
Shave your annual cost-to-serve per user by as little as 10 cents, and that's $100 million that magically appears on the balance sheet.
It's no wonder they obsess about things like building their own switches, servers, operating systems, file systems, etc.
The same incentives don't usually exist in enterprise IT because the scale is much, much smaller, and the portfolio is far more diverse. You're usually providing a very broad range of complex IT services to thousands, not simple web services to billions.
#3 -- Go Look In The Back Room
None of these web-scale companies are exactly forthcoming in this regard, but they *all* run their typical business applications on typical business infrastructure -- at least, to the best of my knowledge.
None of their enterprise IT vendors are usually encouraged to promote this
So, if they have this all-powerful next-generation web-scale cloud, why do they invest in a completely separate stack of technology, people and processes to run their business applications?
Because it makes utter sense -- that's why.
Another View?
At an over-simplified level, a given organization's approach to supporting the business model is driven by three things: the technology model, the financial model and the operational model. All three have to be aligned to support what the business is trying to achieve.
If you consider a web-scale company's cloud, all three are considerably different than the more familiar enterprise IT. Their customer-facing services are built differently, funded differently, and operated differently. It's integral to their business strategy.
The central challenge isn't simply aping what they're doing, it's figuring out your own strategy for your own business model.
Just because we admire a ballet dancer doesn't mean it makes sense for us all to try and become one.
That being said, it's impressive on how these web-scale companies been able to craft their own distinctive approach to powering their businesses with IT.
I find it fascinating -- as do many others -- but not directly relevant.
Yes, inevitably, aspects of what they've innovated will trickle into other more familiar flavors of IT -- when it makes sense, that is.
Uncomfortable truth: enterprise IT is different. Very different.
So let's get back to work, shall we? :)
----------------------
Like this post? Why not subscribe via email?
Comments