You're in the middle of a sequence of posts where I attempt to describe the future state of IT in a short five years from now.
My first post dealt more with existential issues: why does IT exist, how their role is evolving, and what the implications might be.
I ended with a first of many potential interpretations for the term "virtual IT": IT organizations will evolve to an enterprise provider of IT services, "frontsourcing" an ever-growing network of specialized and external providers who can do things better, faster or cheaper.
So, let's keep this discussion moving forward, and dig in to some of the more subtle implications of virtual service providers.
Thriving In the Service Economy
We're all proficient consumers of services in our personal lives. From cable TV to hairdressers, we all have become pretty adept at not only figuring out our own needs, but deciding who we want to give our money to.
Personally, I find myself consuming newer kinds of services as my life situation changes. A while back, I had my first sobering chat with a financial adviser. More recently, I gave up on the notion of mowing my own lawn. More and more home improvement projects are gently steered by my wife to people who actually might know what they're doing.
Not to oversimplify, but when I do this, I'm basically answering three questions:
1 -- do I roughly know what my problem is?
2 -- is this person proficient at understanding my problem, and delivering a good solution?
3 -- do I trust this person?
Fail at any three of these, and there won't be a transaction.
In the course of my work, it's often the case that we need to bring in external specialists on one topic or another. We'll be wrestling with some issue on a concall, and then a new disembodied voice will join the call. After some opening formalities, we can quickly figure out the answers 1,2 and 3 above, and get on to having this person help us. Very often, I don't really know where this person might be (not relevant), or what our overall business relationship might be (managed by someone else at the company), or a bunch of other minor details.
All I know is that we get a quality answer very quickly, under presumably reasonable business terms.
I guess that meets one definition of a "virtual service provider" -- delivering a quality answer quickly under presumably reasonable business terms.
Now, let's map that idealized concept into the world of day-to-day IT.
IT As A (Virtual) Service Provider
The same set of social transaction happens (or should happen) when the business engages IT. And, roughly speaking, the same three questions have to be answered. As a business person, I'm responsible for understanding the problem statement (sometimes termed "requirements", but I find that a very limiting concept).
The (virtual) IT service provider is responsible to establish that they are relatively proficient in understanding my problem, and proposing a solution that's acceptable to me.
And, of course, there's that trust thing ...
If they tell me they're too busy, or that my problem isn't important enough to them, or that I'm basically screwed until the distant future, I'd do what any reasonable consumer of services would do -- keep shopping until I found something that worked.
Does speed of an answer matter? Yes, very often a business user will pay for a good-enough answer today, rather than a theoretically perfect answer in six months. We're not accustomed to expressing the hard value of time-to-solution, but we know it's there.
Can I get a haircut today, or is there a form to fill out, a series of meetings to review my request, a decision to be made by people who don't understand my situation, a detailed negotiation as to who will perform the service, and a likely six month backlog before anyone gets around to it?
If I'm really, really picky about my haircut, maybe I'll put up with all of that. Or, if I just want to stop looking so shaggy, I'll hang up the phone and call someone else. That is, presuming I still had any hair left to cut :-)
Does the quality of service matter? Yes, but it often can be dynamic.
A while back, I was working with contractors on various projects in my house. Some things I really cared about, other things I didn't care so much. What was interesting was that -- as the projects progressed -- my perspectives would often change, sometimes as the result of the contractor (service provider) pointing out one aspect or another than I might have missed.
And the very best contractors were able to flex dynamically with my changing needs over time, and not retreat into a "well, we'll have to get back to you on this".
There was also a whole layer to the house projects that I just didn't want to be bothered with: getting building permits, knowing the building code, getting the work inspected by town officials, selecting and managing other vendors, and so on.
I knew all that stuff was important to keep me out of trouble -- but I expected that my service provider would just take care of those details on my behalf.
I could also describe my contractor as a virtual service provider. I don't really care how and where the resources come from, do I? I'm just looking for an outcome.
If we make this very specific about the near-term future of IT, the analogy is somewhat useful:
- businesses will value a good-enough answer from IT that's delivered quickly
- businesses will value the ability to quickly change their perspectives on what's important, and have IT flexibly respond
- businesses will expect IT to keep them out of trouble on their behalf
- and, of course, to get all of the above on competitive business terms
Now, Compare That Against Traditional Thinking
How much IT thinking is driven by driving out costs to the exclusion of all else? Sure, that makes sense for IT functions that are relatively stable and somewhat uninteresting, but I think business users will come to value speed, flexibility and responsiveness first, safety second, and cost optimization third.
Do we use the processes designed for complex and hairy IT projects to handle the hundreds (or thousands) of minor IT projects that make up the modern landscape?
Do we spend an inordinate amount of time defining static requirements and worst-case scenarios, or are we thinking more along the lines of "let's get started and see where it goes"? I've done enough IT projects where I've gone back and dug up the original planning documents and am always amazed at just how much was wrong, or simply unimportant. And, of course, amazed by all the important stuff we found out along the way that wasn't originally captured.
Does IT view the business as customers and clients? Does the business view IT as one of the best service providers around? Or do all parties tend to fall into a confrontational perspective?
Let's Get Geeky For Just A Moment
My first real exposure to virtualization concepts was way back in the misty times of MVS and BSD UNIX -- and we're talking about virtual memory.
It was a simple concept.
If you're a program, here's an initial set of working pages to get you started, no questions asked. Here are some algorithms that are watching what you're doing, and juggling available resources in a sincere attempt to give you what you want, when you want it, and not hose everyone else. Here are some provisions to keep you -- and everyone else -- out of trouble.
And, oh by the way, you'll end up with a better way of using an expensive resource -- main memory -- than if you tried to do hand-craft a "better" solution.
Will Virtual IT Be That Much Different?
Here's something you'd like to get done.
I'll give you an initial set of resources to get you started quickly, no questions asked.
I'll put some algorithms in place to watch what you're doing, and automatically feed in more (or less) resource depending on what you might need over time.
I'll put some safeguards in place to keep you and the rest of us out of trouble.
And, oh by the way, we'll all end up with a better way of using expensive IT resources (hardware, software and people) than if we tried to hand-craft every little piece along the way.
And, so, we have a second possible interpretation (among many) of "virtual IT", don't we?
And, in a certain sense, haven't we seen some of it before?
Courteous comments always welcome ...
Comments