Competition is the engine that drives us all forward.
At an individual level, we compete with others who might be able to do a better job, or do things for a lower cost. At a corporate level, our companies compete each and every day for customers.
Organizations within companies have learned to compete as well. If you deliver a service to the business, there are always external competitors who'd like to show your internal customers a better, faster, cheaper way to get things done.
I believe the necessity to compete for internal customers is the primary engine behind IT-as-a-service concepts.
Fail to successfully compete, and your internal customers will likely go outside to get what they need.
In this post, I'd like to share with the the most cogent explanation I've seen on this topic to date: what's happening, what it means to IT organizations everywhere, and -- most importantly -- what you can do about it.
At an individual level, we compete with others who might be able to do a better job, or do things for a lower cost. At a corporate level, our companies compete each and every day for customers.
Organizations within companies have learned to compete as well. If you deliver a service to the business, there are always external competitors who'd like to show your internal customers a better, faster, cheaper way to get things done.
I believe the necessity to compete for internal customers is the primary engine behind IT-as-a-service concepts.
Fail to successfully compete, and your internal customers will likely go outside to get what they need.
In this post, I'd like to share with the the most cogent explanation I've seen on this topic to date: what's happening, what it means to IT organizations everywhere, and -- most importantly -- what you can do about it.
Welcome To The New IT Department
The best way to get into this discussion is to take the perspective of the business person who'd like to get something important done.
Maybe it's a new requirement, maybe it's a new idea, maybe it's a new marketplace -- it doesn't matter.
As a business person at EMC, I often start presentations by introducing my new IT department -- it's my corporate credit card. I can potentially consume a *lot* of IT externally without having to really answer to anyone if I choose.
These external providers offer attractive services, and let me select and consume what I need with a minimum of length approvals, negotiations, project requests, planning meetings, etc. Bottom line: if I can find what I need externally, I win. I can get where I'm going faster.
Now, in all fairness, there are policies and practices in place that are designed to keep me from doing unwise things, but I've got considerable lattitude nonetheless. Especially if I can get something going before anyone finds out.
The truth behind this phenomenon is simple: we now have a generation of knowledge workers, managers and business leaders that are quite comfortable with technology. They use it every day, and they know what's out there. Personal information technology isn't the mystery it was ten or twenty years ago.
When I hear people use the phrase "consumerization of IT", it isn't about the technology getting dumber; it's about the IT consumers getting smarter -- and more demanding.
Where Will Your Business Source IT Requirements?
I know what many of the IT professionals reading this might be thinking: "well, we have policies in place to keep that sort of thing from happening".
I'm sure you do, but -- frankly -- I think you're underestimating the persistence of a motivated business worker, manager or leader.
Collectively, we're trained and encouraged to solve problems and find creative solutions -- and if that means going outside of enterprise IT, so be it. Better to ask forgiveness than ask permission.
As a popular example, it's an infrequent case when Salesforce.com's CRM application is brought forward by the IT department as a "better solution". Or Google Mail. Or Dropbox. Or iPads, for that matter :)
The bottom line? People have more choices than ever -- and there's no turning back.
Enterprise IT Customers Are Looking Elsewhere
Need evidence? There's plenty to be found -- if not in your immediate range, then certainly from any number of authoritative external sources.
Even though the numbers shown here are substantial, I think they're understated in several regards.
The more of a knowledge worker you are, the more likely you're using external IT services and devices to get your job done. As business models shift rapidly in favor of these knowledge workers, you'll likely see even higher rates of unauthorized IT usage.
Besides, who would admit to a survey taker that they're doing something, err, unapproved?
The prognosis is clear: every year, you'll see more external services, better external services -- and better marketed external services. There are untold billions being invested in external service provider models, and they're targeting the people who can use their services -- sometimes with the cooperation of IT, sometimes not.
Something to look forward to: just imagine what happens when all those corporate Apple users get their hands on iCloud ...
IT Organizations Must Now Learn To Compete
I often get asked by journalists and analysts as to what might be driving all this interest in IT-as-a-service.
It's simple: IT organizations now have to learn to compete for their internal customers, just like other internal corporate support functions.
Fail to learn to compete, and the future is not pretty.
For starters, IT can lose control -- if they haven't already. Whether the "shadow IT" is showing up in desktops and closets, or at any number of external service providers, people inevitably find a way to get their jobs done.
Continue to lose control, and internal IT inevitably becomes less relevant to the people they serve. Becoming less relevant inevitably results in budget cuts, staffing cuts, or often both. That's how the world works.
If the organization still believes internal IT is important to the overall strategy, they'll invest in new IT leadership. At various events, I often ask the question "how many of you have had a significant change in IT leadership in the last 24 months?".
Every time, about half the hands go up.
If, on the other hand, the business doesn't see IT having the potential to deliver any unique value, that's when cost-motivated outsourcing becomes popular. Why are we doing this ourselves when others can do it cheaper -- and maybe better?
How IT Service Providers Compete
Perhaps the most useful example here is to look more closely at organizations that have to compete for IT customers each and every day -- and that's the current crop of enterprise IT service providers.
They have no monopoly. They have to earn their customers' loyalty each and every day.
For starters, the mindset is usually different -- they partner with the people using their services. They invest the time and cycles to really understand what their customers are trying to get done each and every day. Their ultimate goal is a relationship, not a quick sale.
Around that relationship, they'll try and offer as many related services as possible. This works for the service provider, because the costs associated with acquiring and maintaining a customer relationship are substantial -- best to sell those people as many related things as possible. This also works for their customers, because -- all things being equal -- it's easier to work with fewer external providers who can offer more breadth of capability.
These same service providers will also invest in something that internal IT organizations almost never consider: a sales and marketing function. Now, I'm not suggesting that every enterprise IT organization go hire a sales force with quota and a full-time marketing department, but there ought to be at least some semblance of those skills, behaviors and roles in place.
Part of the attractiveness of working with external providers is that they're trying to earn your business. Part of the frustration that many business users of IT share with me is that they don't see this from their internal IT functions.
Needless to say, all pricing in any service provider model is completely transparent -- you know what you're paying for. And, if you look in their data centers, you'll tend to find IT platforms and operations designed to deliver as many services as possible from a single "factory" -- with an emphasis on simplicity, efficiency -- and trust.
Being A Competitive Internal IT Service Provider Is Different
Having spent much time in both traditional enterprise IT environments and their service provider counterparts, there are some stark differences at a fundamental level.
For one thing, consumption of IT services is designed to be as easy as possible for the people using the services. External IT service providers aren't in the business of rationing IT consumption as so many enterprise IT groups feel they have to do.
The operational model is also different: it's all about the services stack: which services, how popular, how reliable, how flexible, how efficient, etc. To be absolutely clear, I'm not just talking about the services that the customer sees; I'm talking about services that are used by others in the IT operation to service customers.
Ultimately, the technology model is different: it's all about having the minimum number of platforms that can deliver the widest range of shared services. Everything virtualized is only table stakes; there's a strong motivation to use the same platforms over and over again vs. creating new for each and every new service offered.
And There's More
Digging deeper, there are even more fundamental differences between competitive IT service providers and their enteprise IT equivalents.
As mentioned before, IT service providers invest in relationships with the people consuming their services. Not a "well, we had a planning meeting" relationships, but taking the time to know what people are trying to get done -- and how the service provider can position themselves to be seen as part of the solution vs. part of the problem.
IT service providers tend to have have different relationships with their IT suppliers as well: they often bet big on a few key suppliers vs. compete each and every new requirement. Remember, these folks are highly motivated to standardize their
platforms and operations; so the less diversity and tighter integration, the better.
Finally, the most intriguing observation is that their organizational models are very different than their enterprise IT counterparts. The model is built around delivering attractive and efficient IT services vs. one-off large and highly customized IT projects.
Becoming A Competitive Internal IT Service Provider
Not everything directly translates between these two worlds. But enough of the thinking *does* translate that the external IT service provider model is worthy of serious consideration as a starting point.
In this model, enterprise IT becomes a provider of attractive and competitive IT services. There's complete transparency as to what the options are, and what they cost the consumer.
Some of those services might be provided with internal resources. Others might be sourced externally. And the mix between the two will most certainly change over time.
Internal customers now can turn to internal IT for the same reason they often go external: agility -- the ability to move faster and change direction more nimbly.
At the same time, internal IT has the mechanisms in place to maintain control and compliance -- and, in this model, those attributes are built into the internal services model vs. enforced separately.
IT invests in creating the absolute minimum number of platforms to deliver the maximum breadth of required services -- no more building one-offs for each and every new requirement in the enterprise.
Ultimately, internal IT organizations can play the one card that no external service provider can: they have the ability to understand more about their customers' requirements -- and the business they serve -- than most any external provider.
Put differently, transitioning to an internal IT service provider model helps IT focus on what they're *uniquely* qualified to do best. And that attribute is the long-term secret to delivering value vs. being seen solely as a cost-center.
IT Transformation Enablers and Components
So, let's circle back and dig in a bit more with the key elements involved in a transition to an internal IT service provider model.
The first logical bucket to consider is the consumption model.
For starters, it's creating a layered self-service catalog that's relevant, attractive and easy to consume. Keep in mind, most ITaaS services are often consumed by other parts of the IT organization or authorized non-IT users vs. ordinary end users.
Costs (actually prices) need to be associated with each service offered. Indeed, getting to the true cost of delivering a service is often no easy task in itself. Since costs and prices aren't the same thing, there's a second round of thinking on how best to price services to drive the consumption behavior you're looking for.
Finally, like any retailer (or restaurant!) you have to have enough capacity on hand to meet the immediate needs of your customers. Making people wait excessively isn't good for business :)
A second logical bucket is the technology model.
Lots to talk about here (especially if you're a technologist), but -- at a high level, we've got a cloud model for internal IT (e.g. private cloud), a cloud model for external services (e.g. public cloud), and the ultimate goal of creating a rationalized ballance between the two (e.g. hybrid cloud).
Beneath that is the need for IT to remain in control. "Trust" in the context is more than just security or compliance -- it includes things like service levels delivered, availability and recoverability -- all the things that go into someone "trusting" a service and its provider.
The phrase used here ("converged infrastructure") really doesn't do the broader platform discussion any justice. We're talking about biasing towards a dynamically shared pool of resources vs. building little individual technology puddles for each and every requirement. The "converged" descriptor does do a better job of characterizing the operational models used here.
And, finally, the third logical bucket is the IT organization itself: the skills portfolio, and how that's organized into roles and associated processes.
More On Consumption Models
Going even deeper, there's more that can be said about consumption models.
For starters, the big "aha" over the last year is that the primary consumer (and beneficiary) of IT service catalogs is -- well -- IT itself.
Example: many IT projects require some form of infrastructure: some transient, some permanent. Why not start by making your infrastructure-as-a-service cater to those internal-to-IT needs? Same for platform as a service, etc.
Even if your interface to non-IT users is more traditional or consultative, those new requirements can be met far faster and far more efficienctly if they're built on top of an agile catalog of IT services, ready to consume.
Many of us think that the real goal of showback and chargeback isn't so much about covering costs; it's about helping individuals, groups and the overall organization to make intelligent choices around IT consumption. That's one area where there's a clear departure between external IT service providers and their aspiring internal IT equivalents.
Just-in-time capacity isn't solely about being able to serve your internal customers' needs more quickly; it's about pooling and sharing all sorts of resources (equipment, software licenses, skills, etc.) and standardizing the ingredients.
I've seen both service providers and the more progressive internal IT equivalents get very smart on sourcing technology when they can lock in on fewer moving pieces that are broadly used.
New ITaaS Models Based On Cloud
Since we at EMC seem to spend a lot of time on infrastructure, we often fall into the trap of characterizing "cloud stacks" as mostly infrastructure and the software required to manage, orchestrate and secure them.
Taking a slightly broader perspective, there's much more in play here.
There's the need to "cloudify" large and complex enterprise applications that won't be rewritten anytime soon. That means being able to containerize them and enable them to run more efficiently (whether internally or externally), and is an important discussion in its own right -- and we'll be talking more about the VCUBE approach to this starting next year.
In parallel, new applications are being created all the time. Indeed, VMware's broader strategy here around Spring and CloudFoundry represents a new model for creating enterprise-class, cloud-friendly applications from the outset.
And, every day, there are ever more pure SaaS offerings that meet the needs of the business better than any internal effort could hope to achieve. Figuring out how to evolve the entire application portfolio forward in an IT-as-a-Service context is no small feat in itself.
I think it's no surprise to anyone that we're witnessing the early days of a revolution on the client side of information access. The familiar desktop paradigm is now joined by finger-friendly applications running on various mobile devices.
Simply re-creating a full-screen web or desktop experience isn't exactly what people want going forward. Going forward, IT will be responsible for the experience, and not necessarily the device.
From an enabling technology perspective, EMC has focused on three areas that we think will be key in this world to extend what virtualization is already doing.
The first is creating "trust" with various cloud models -- more than just security, it's being able to assure service delivery regardless of the circumstances. The second is information mobility -- in this
world, information will be moving around a lot, and ordinary replication or data transfer mechanisms won't fill the bill.
And the third is orchestration and automation -- the ability to manage processes and outcomes vs. individual tasks.
Philisophically, we strongly believe these capabilities should be built into the services themselves, and transparently available for any and all use cases that come along.
New Organizational Models -- From Silos To Services
If you're a regular reader of this blog, I've talked about this at length -- and here is another representation of the same ideas, perhaps with better formatting :)
The core idea here is simple: if you're going to be in the business of delivering attractive and competitive services, you're going to have to organize around that principle.
In this model, there's less emphasis on the specific technical disciplines, and more emphasis in creating and delivering services.
The model here shows four: infrastructure, app platform (database plus other components), application and user experience. Each service category typically consumes services from beneath it, e.g. user experience service consumes app
services, app services consume app platform services, and everything consumes infrastructure.
By "platform", the intention is to represent "the platform that delivers those unique services", e.g. app support environment platform, specific application platforms, specific user experience platform and so on.
And infrastructure, of course ...
The emphasis on "IT service management" and "IT services marketing" is intentional. We've chosen to over-emphasize the marketing aspect for one simple reason: it's almost always missing from IT organizational DNA, and that lack ends up hurting badly in this model.
Focus On The Business
If we dig in deeper to what goes on here in these two disciplines, you'll see familiar activities, but completely oriented around what people actually want, vs. what IT decides to build.
I call this shift in mindset the "retail gene".
You'll find it all around you -- except in many IT organizations. What do people want and need? How do they want to consume it? How do we productize it? How do we market it? How do we measure it? How do we support and manage it?
Just like any MBA studying a market :)
Indeed, I'd argue that the required skills here aren't all that hard to find -- once you step out of the rather cloistered world of IT professionals.
The EMC IT Example
As many of you know, we often use our own IT organization's transformation as an illustrative example of what's involved. Sure, every situation is different, but we've seen enough similarities between our own experiences and those of our customers that we think the patterns and models here are very useful.
One of the segments that customers usually find quite interesting is the detail behind how our infrastructure organizational model shifted dramatically between 2008 and the present-state 2011. Had we known what we do now, we could have gotten there far faster and with a lot less churn -- one of the reasons we're highly motivated to share our experiences and make it easier for others.
The starting point was the traditional technology silos that we're all familiar with. Progressively, the organization morphed into a services-oriented structure where the definition, promotion and delivery of services now is the organizing principle.
The same model is now being progressively applied to app platforms, enterprise applications, as well as user experiences.
Quantifying The Benefits
Frequently, I get asked -- how do you measure success when considering a transformation like this?
I think it's a fair question -- even if you "get it", there's the unenviable task of having to convince many others.
Although EMC offers popular consulting services to help make the case for a transformation investment (and quantify resulting benefits), it's often useful to share our own internal EMC IT experiences.
The chart here graphically shows our progress, using "degree of virtualization" as a proxy for how much we've transitioned to an IT-as-a-service model. Yes, the "percent virtualized" metric isn't perfect in several regards, but it is widely comparable across organizations.
Right now, we're stating that we're around 80 to 85% virtualized across our landscape. Some smaller portion of that runs in the IT-as-a-service model, but the gap between "virtualized" and run as-a-service is closing fast.
We can do a good job of quantifying the benefits at each stage of the journey (both "hard" and "soft" metrics), but -- here's the important point -- the ideal measurements change from stage to stage.
In the first phase where you're simply virtualizing non-critical IT-owned applications, it's all about cost optimization: spending less on servers, storage, etc. and the people who have to administer them. Most everyone understands what simple virtualization can do in this regard.
In the second phase, it's more about delivering better IT services under the traditional guise but using virtualization as a foundation. By better IT, I mean "fewer outages" and "better data protection" and "more consistent IT compliance" and "reduced operational effort".
As well as saving some serious money :)
In the third phase, it's really all about being more responsive to the business -- agility -- as well as being able to deliver better IT services (more robust, more secure, more flexible) and, of course, even more cost savings.
One Of The More Interesting Charts
About a month ago, the EMC IT folks started using this chart, showing how -- as agility increased -- the IT spend dramatically shifted from "keeping the lights on" to "investing in new capabilities".
This outcome shouldn't be surprising -- as the back end of IT gets more nimble and efficient at delivering easy-to-consume and flexible services, there's more time, energy and resources in doing the new stuff. And those new and innovative capabilities are always big consumers of the service catalog -- as you'd expect.
As an exercise for the reader, where would you put your current organization on time to provision a developer-ready or user-ready application environment? Not just the server -- everything you need for it to be directly usable for the intended
purpose.
Or, in terms of how much of the IT spend goes to doing new initiatives vs. feeding as-is operations?
Just knowing what the answers might be is a useful exercise.
EMC IT's Cloud 9 Example
Here's just one of the internal services being offered by our internal IT team, dubbed "Cloud 9". Note: a little marketing can go a long way :)
It's self-service infrastructure for anyone who wants it, typically for transient requirements of 90 days or less.
Longer use and/or more demanding requirements initiates a separate process, but if you can stay within the confines of what the service offers, you get what you need -- no questions asked.
The biggest audience? Internal software developers -- both within EMC IT and our various product development communities.
What you can't see from this slide is what's behind it: a small team of people talking to the users of the service, figuring out how they're using it, discovering unmet needs, trying to improve the service, etc.
Put differently, this wasn't your typical hit-and-run IT project; it's part of a sustaining organizational function to deliver internal IT services that people want to consume.
And that makes all the difference in the end.
Being Public About Inhibitors
Change isn't easy, and we're doing everything we can to share not only our internal challenges, but the challenges we hear from our customers who are doing the same. Here's a summarized list.
On the technology front, no real surprises here.
Building and running large, shared infrastructures that deliver a variety of different services is very different than the one-application, one-infrastructure approach we're all so familiar with.
It's not that it's intrinsically hard; it's just different.
As you increase the degree of automation, the people responsible for the outcomes get progressively more nervous. You need to allow time for comfort and experience to develop around a certain degree of automation before progressing to the next level.
Again, it's really human nature complemented by practical experience.
On the operations front, there's a lot to address. You're basically building a new function that perhaps didn't exist before. There are new management processes, new life-cycle processes, and all the skills, roles and org structures that go with it.
The challenge here seems to be directly proportional to the size of the IT operation, with smaller IT organizations having a decided advantage.
The real heavy lifting happens at the interface between IT and the business. There's the need to establish a service-oriented culture when non-IT people interact with IT people. There's the need to teach businesses to consume off the service catalog vs. buying stuff and handing it over to IT to run.
And finance has to get involved to not only support the shared services model, but also take responsibility for managing overall levels of IT spend -- in this model, IT can't be doing that and be expected to be successful.
Advice On Getting Started
Again, every situation is different, but we've seen enough examples where some useful patterns are starting to emerge.
First, there needs to be a clear case made for investing in a transformation.
Going from one style of IT to another isn't simple, easy or cheap. The case needs to be made to executive management, business unit leaders, and the IT organization itself.
More than a few IT leaders have launched themselves on a transformational journey without taking the time for this important step. As a result, they've had to stop, go back, and retrace their path before proceeding again. Or, they've misjudged their readiness to move ahead -- and things end up taking much longer than they should. Either way, a key step in the journey for most.
A second key step is finding business champions outside of IT -- people with a vested stake in consuming the services to be delivered. Not so much as a source of funding, but as your first "target customers" who can validate and consume what you intend to deliver.
A third key step is to break off a small team from the mainstream IT organization, and empower them to move ahead quickly without having to drag the legacy behind them. As they mature their capabilities and processes, they can progressively handle more and more of the IT workloads using the new ITaaS model.
A fourth key step is to design the first round of services for other IT groups who can benefit from attractice and easy to consume services. Almost every time, this turns out to be infrastructure services for application development and testing. This "do it inside of IT first" gets to a quick and visible win without having to change the relationship between IT and its users at the outset.
Finally, be very open and honest about the inhibitors along the way. Change isn't easy, and being absolutely transparent about the challenges ahead makes them a bit less, well, challenging. That's worked for us internally here at EMC; we'd highly recommend it for others as well.
How EMC Can Help
Since we've been working on this for a while, we've amassed a considerable portfolio of expertise, solutions and foundational technologies that can greatly accelerate our customers' journey to a competitive ITaaS model.
We can't cover everything we and our partners do here, so consider this a sampling of highlights.
In addition to the "making the case" and "readiness assessment" consulting work described above, we can dive down deeper into the next layer of tasks. We can help IT organizations construct their first service catalogs, do the consumption modeling, and help get a handle on the financial model as well.
We also have a healthy and growing roster of compatible service providers, which play an important role. Not only can you benefit from understanding from what they're doing (example: learn from their rate cards!), they're an important consumption option that can help accelerate the transition.
More than a few of our customers have started their journey with rented infrastructure and capabilities from a compatible service provider, enabling a quick win to the business while building the case for a more substantial internal investment. They can focus on making the front-end of IT-as-a-service work without having to invest in the back end up front. When it's time, the move is easy. Or the workloads can stay where they are if you're happy :)
When it comes to operations, there's a lot to be done, and we're prepared to help.
We can identify the key skills and roles you're going to need, and various options for sourcing the talent. We can define the operational process you'll need, how you'll get there, and how you'll be continually improving them. We can show you organizational models that work, and help you plan on how to make them work in your own organization.
Finally, there's technology to consider.
From Vblocks to data protection to security to our EMC Proven Solutions; the technology you'll need is there, and it's relatively easy to deploy and operate -- whether you do it yourself, or use our implementation and migration services.
Final Thoughts
Going from a world where you don't have to compete -- to one where you do have to compete -- can be jarring to everyone involved. Whether you're an IT specialist, IT manager or IT leader, it's a completely different game when your capabilities are rigorously stacked up against those of others.
Part of the inherent challenge is simply recognizing what's happening, and starting to do something about it. There are literally hundreds of external IT service providers that compete for enterprise IT business each and every day, and are successful at it. And there's no reason why an internal IT function couldn't do the same -- but only better, due to the natural "home field" advantage.
IT leadership, in particular, is finding themselves especially challenged. The model for IT success is shifting rapidly -- it's now about delivering valuable IT services that the business wants and needs, doing so at a reasonable cost, while maintaining the control that the organization demands.
The good news is that we now have plenty of examples of IT organizations that made the change, and are starting to see the benefits. It's working for them. And those that have made the decision to invest in transformation are justifiably proud of what they and their teams have achieved.
And, make no mistake, we're proud at EMC to be a part of their success.
Who said IT was boring?
By: Chuck Hollis
The best way to get into this discussion is to take the perspective of the business person who'd like to get something important done.
Maybe it's a new requirement, maybe it's a new idea, maybe it's a new marketplace -- it doesn't matter.
As a business person at EMC, I often start presentations by introducing my new IT department -- it's my corporate credit card. I can potentially consume a *lot* of IT externally without having to really answer to anyone if I choose.
These external providers offer attractive services, and let me select and consume what I need with a minimum of length approvals, negotiations, project requests, planning meetings, etc. Bottom line: if I can find what I need externally, I win. I can get where I'm going faster.
Now, in all fairness, there are policies and practices in place that are designed to keep me from doing unwise things, but I've got considerable lattitude nonetheless. Especially if I can get something going before anyone finds out.
The truth behind this phenomenon is simple: we now have a generation of knowledge workers, managers and business leaders that are quite comfortable with technology. They use it every day, and they know what's out there. Personal information technology isn't the mystery it was ten or twenty years ago.
When I hear people use the phrase "consumerization of IT", it isn't about the technology getting dumber; it's about the IT consumers getting smarter -- and more demanding.
Where Will Your Business Source IT Requirements?
I know what many of the IT professionals reading this might be thinking: "well, we have policies in place to keep that sort of thing from happening".
I'm sure you do, but -- frankly -- I think you're underestimating the persistence of a motivated business worker, manager or leader.
Collectively, we're trained and encouraged to solve problems and find creative solutions -- and if that means going outside of enterprise IT, so be it. Better to ask forgiveness than ask permission.
As a popular example, it's an infrequent case when Salesforce.com's CRM application is brought forward by the IT department as a "better solution". Or Google Mail. Or Dropbox. Or iPads, for that matter :)
The bottom line? People have more choices than ever -- and there's no turning back.
Enterprise IT Customers Are Looking Elsewhere
Need evidence? There's plenty to be found -- if not in your immediate range, then certainly from any number of authoritative external sources.
Even though the numbers shown here are substantial, I think they're understated in several regards.
The more of a knowledge worker you are, the more likely you're using external IT services and devices to get your job done. As business models shift rapidly in favor of these knowledge workers, you'll likely see even higher rates of unauthorized IT usage.
Besides, who would admit to a survey taker that they're doing something, err, unapproved?
The prognosis is clear: every year, you'll see more external services, better external services -- and better marketed external services. There are untold billions being invested in external service provider models, and they're targeting the people who can use their services -- sometimes with the cooperation of IT, sometimes not.
Something to look forward to: just imagine what happens when all those corporate Apple users get their hands on iCloud ...
IT Organizations Must Now Learn To Compete
I often get asked by journalists and analysts as to what might be driving all this interest in IT-as-a-service.
It's simple: IT organizations now have to learn to compete for their internal customers, just like other internal corporate support functions.
Fail to learn to compete, and the future is not pretty.
For starters, IT can lose control -- if they haven't already. Whether the "shadow IT" is showing up in desktops and closets, or at any number of external service providers, people inevitably find a way to get their jobs done.
Continue to lose control, and internal IT inevitably becomes less relevant to the people they serve. Becoming less relevant inevitably results in budget cuts, staffing cuts, or often both. That's how the world works.
If the organization still believes internal IT is important to the overall strategy, they'll invest in new IT leadership. At various events, I often ask the question "how many of you have had a significant change in IT leadership in the last 24 months?".
Every time, about half the hands go up.
If, on the other hand, the business doesn't see IT having the potential to deliver any unique value, that's when cost-motivated outsourcing becomes popular. Why are we doing this ourselves when others can do it cheaper -- and maybe better?
How IT Service Providers Compete
Perhaps the most useful example here is to look more closely at organizations that have to compete for IT customers each and every day -- and that's the current crop of enterprise IT service providers.
They have no monopoly. They have to earn their customers' loyalty each and every day.
For starters, the mindset is usually different -- they partner with the people using their services. They invest the time and cycles to really understand what their customers are trying to get done each and every day. Their ultimate goal is a relationship, not a quick sale.
Around that relationship, they'll try and offer as many related services as possible. This works for the service provider, because the costs associated with acquiring and maintaining a customer relationship are substantial -- best to sell those people as many related things as possible. This also works for their customers, because -- all things being equal -- it's easier to work with fewer external providers who can offer more breadth of capability.
These same service providers will also invest in something that internal IT organizations almost never consider: a sales and marketing function. Now, I'm not suggesting that every enterprise IT organization go hire a sales force with quota and a full-time marketing department, but there ought to be at least some semblance of those skills, behaviors and roles in place.
Part of the attractiveness of working with external providers is that they're trying to earn your business. Part of the frustration that many business users of IT share with me is that they don't see this from their internal IT functions.
Needless to say, all pricing in any service provider model is completely transparent -- you know what you're paying for. And, if you look in their data centers, you'll tend to find IT platforms and operations designed to deliver as many services as possible from a single "factory" -- with an emphasis on simplicity, efficiency -- and trust.
Being A Competitive Internal IT Service Provider Is Different
Having spent much time in both traditional enterprise IT environments and their service provider counterparts, there are some stark differences at a fundamental level.
For one thing, consumption of IT services is designed to be as easy as possible for the people using the services. External IT service providers aren't in the business of rationing IT consumption as so many enterprise IT groups feel they have to do.
The operational model is also different: it's all about the services stack: which services, how popular, how reliable, how flexible, how efficient, etc. To be absolutely clear, I'm not just talking about the services that the customer sees; I'm talking about services that are used by others in the IT operation to service customers.
Ultimately, the technology model is different: it's all about having the minimum number of platforms that can deliver the widest range of shared services. Everything virtualized is only table stakes; there's a strong motivation to use the same platforms over and over again vs. creating new for each and every new service offered.
And There's More
Digging deeper, there are even more fundamental differences between competitive IT service providers and their enteprise IT equivalents.
As mentioned before, IT service providers invest in relationships with the people consuming their services. Not a "well, we had a planning meeting" relationships, but taking the time to know what people are trying to get done -- and how the service provider can position themselves to be seen as part of the solution vs. part of the problem.
IT service providers tend to have have different relationships with their IT suppliers as well: they often bet big on a few key suppliers vs. compete each and every new requirement. Remember, these folks are highly motivated to standardize their
platforms and operations; so the less diversity and tighter integration, the better.
Finally, the most intriguing observation is that their organizational models are very different than their enterprise IT counterparts. The model is built around delivering attractive and efficient IT services vs. one-off large and highly customized IT projects.
Becoming A Competitive Internal IT Service Provider
Not everything directly translates between these two worlds. But enough of the thinking *does* translate that the external IT service provider model is worthy of serious consideration as a starting point.
In this model, enterprise IT becomes a provider of attractive and competitive IT services. There's complete transparency as to what the options are, and what they cost the consumer.
Some of those services might be provided with internal resources. Others might be sourced externally. And the mix between the two will most certainly change over time.
Internal customers now can turn to internal IT for the same reason they often go external: agility -- the ability to move faster and change direction more nimbly.
At the same time, internal IT has the mechanisms in place to maintain control and compliance -- and, in this model, those attributes are built into the internal services model vs. enforced separately.
IT invests in creating the absolute minimum number of platforms to deliver the maximum breadth of required services -- no more building one-offs for each and every new requirement in the enterprise.
Ultimately, internal IT organizations can play the one card that no external service provider can: they have the ability to understand more about their customers' requirements -- and the business they serve -- than most any external provider.
Put differently, transitioning to an internal IT service provider model helps IT focus on what they're *uniquely* qualified to do best. And that attribute is the long-term secret to delivering value vs. being seen solely as a cost-center.
IT Transformation Enablers and Components
So, let's circle back and dig in a bit more with the key elements involved in a transition to an internal IT service provider model.
The first logical bucket to consider is the consumption model.
For starters, it's creating a layered self-service catalog that's relevant, attractive and easy to consume. Keep in mind, most ITaaS services are often consumed by other parts of the IT organization or authorized non-IT users vs. ordinary end users.
Costs (actually prices) need to be associated with each service offered. Indeed, getting to the true cost of delivering a service is often no easy task in itself. Since costs and prices aren't the same thing, there's a second round of thinking on how best to price services to drive the consumption behavior you're looking for.
Finally, like any retailer (or restaurant!) you have to have enough capacity on hand to meet the immediate needs of your customers. Making people wait excessively isn't good for business :)
A second logical bucket is the technology model.
Lots to talk about here (especially if you're a technologist), but -- at a high level, we've got a cloud model for internal IT (e.g. private cloud), a cloud model for external services (e.g. public cloud), and the ultimate goal of creating a rationalized ballance between the two (e.g. hybrid cloud).
Beneath that is the need for IT to remain in control. "Trust" in the context is more than just security or compliance -- it includes things like service levels delivered, availability and recoverability -- all the things that go into someone "trusting" a service and its provider.
The phrase used here ("converged infrastructure") really doesn't do the broader platform discussion any justice. We're talking about biasing towards a dynamically shared pool of resources vs. building little individual technology puddles for each and every requirement. The "converged" descriptor does do a better job of characterizing the operational models used here.
And, finally, the third logical bucket is the IT organization itself: the skills portfolio, and how that's organized into roles and associated processes.
More On Consumption Models
Going even deeper, there's more that can be said about consumption models.
For starters, the big "aha" over the last year is that the primary consumer (and beneficiary) of IT service catalogs is -- well -- IT itself.
Example: many IT projects require some form of infrastructure: some transient, some permanent. Why not start by making your infrastructure-as-a-service cater to those internal-to-IT needs? Same for platform as a service, etc.
Even if your interface to non-IT users is more traditional or consultative, those new requirements can be met far faster and far more efficienctly if they're built on top of an agile catalog of IT services, ready to consume.
Many of us think that the real goal of showback and chargeback isn't so much about covering costs; it's about helping individuals, groups and the overall organization to make intelligent choices around IT consumption. That's one area where there's a clear departure between external IT service providers and their aspiring internal IT equivalents.
Just-in-time capacity isn't solely about being able to serve your internal customers' needs more quickly; it's about pooling and sharing all sorts of resources (equipment, software licenses, skills, etc.) and standardizing the ingredients.
I've seen both service providers and the more progressive internal IT equivalents get very smart on sourcing technology when they can lock in on fewer moving pieces that are broadly used.
New ITaaS Models Based On Cloud
Since we at EMC seem to spend a lot of time on infrastructure, we often fall into the trap of characterizing "cloud stacks" as mostly infrastructure and the software required to manage, orchestrate and secure them.
Taking a slightly broader perspective, there's much more in play here.
There's the need to "cloudify" large and complex enterprise applications that won't be rewritten anytime soon. That means being able to containerize them and enable them to run more efficiently (whether internally or externally), and is an important discussion in its own right -- and we'll be talking more about the VCUBE approach to this starting next year.
In parallel, new applications are being created all the time. Indeed, VMware's broader strategy here around Spring and CloudFoundry represents a new model for creating enterprise-class, cloud-friendly applications from the outset.
And, every day, there are ever more pure SaaS offerings that meet the needs of the business better than any internal effort could hope to achieve. Figuring out how to evolve the entire application portfolio forward in an IT-as-a-Service context is no small feat in itself.
I think it's no surprise to anyone that we're witnessing the early days of a revolution on the client side of information access. The familiar desktop paradigm is now joined by finger-friendly applications running on various mobile devices.
Simply re-creating a full-screen web or desktop experience isn't exactly what people want going forward. Going forward, IT will be responsible for the experience, and not necessarily the device.
From an enabling technology perspective, EMC has focused on three areas that we think will be key in this world to extend what virtualization is already doing.
The first is creating "trust" with various cloud models -- more than just security, it's being able to assure service delivery regardless of the circumstances. The second is information mobility -- in this
world, information will be moving around a lot, and ordinary replication or data transfer mechanisms won't fill the bill.
And the third is orchestration and automation -- the ability to manage processes and outcomes vs. individual tasks.
Philisophically, we strongly believe these capabilities should be built into the services themselves, and transparently available for any and all use cases that come along.
New Organizational Models -- From Silos To Services
If you're a regular reader of this blog, I've talked about this at length -- and here is another representation of the same ideas, perhaps with better formatting :)
The core idea here is simple: if you're going to be in the business of delivering attractive and competitive services, you're going to have to organize around that principle.
In this model, there's less emphasis on the specific technical disciplines, and more emphasis in creating and delivering services.
The model here shows four: infrastructure, app platform (database plus other components), application and user experience. Each service category typically consumes services from beneath it, e.g. user experience service consumes app
services, app services consume app platform services, and everything consumes infrastructure.
By "platform", the intention is to represent "the platform that delivers those unique services", e.g. app support environment platform, specific application platforms, specific user experience platform and so on.
And infrastructure, of course ...
The emphasis on "IT service management" and "IT services marketing" is intentional. We've chosen to over-emphasize the marketing aspect for one simple reason: it's almost always missing from IT organizational DNA, and that lack ends up hurting badly in this model.
Focus On The Business
If we dig in deeper to what goes on here in these two disciplines, you'll see familiar activities, but completely oriented around what people actually want, vs. what IT decides to build.
I call this shift in mindset the "retail gene".
You'll find it all around you -- except in many IT organizations. What do people want and need? How do they want to consume it? How do we productize it? How do we market it? How do we measure it? How do we support and manage it?
Just like any MBA studying a market :)
Indeed, I'd argue that the required skills here aren't all that hard to find -- once you step out of the rather cloistered world of IT professionals.
The EMC IT Example
As many of you know, we often use our own IT organization's transformation as an illustrative example of what's involved. Sure, every situation is different, but we've seen enough similarities between our own experiences and those of our customers that we think the patterns and models here are very useful.
One of the segments that customers usually find quite interesting is the detail behind how our infrastructure organizational model shifted dramatically between 2008 and the present-state 2011. Had we known what we do now, we could have gotten there far faster and with a lot less churn -- one of the reasons we're highly motivated to share our experiences and make it easier for others.
The starting point was the traditional technology silos that we're all familiar with. Progressively, the organization morphed into a services-oriented structure where the definition, promotion and delivery of services now is the organizing principle.
The same model is now being progressively applied to app platforms, enterprise applications, as well as user experiences.
Quantifying The Benefits
Frequently, I get asked -- how do you measure success when considering a transformation like this?
I think it's a fair question -- even if you "get it", there's the unenviable task of having to convince many others.
Although EMC offers popular consulting services to help make the case for a transformation investment (and quantify resulting benefits), it's often useful to share our own internal EMC IT experiences.
The chart here graphically shows our progress, using "degree of virtualization" as a proxy for how much we've transitioned to an IT-as-a-service model. Yes, the "percent virtualized" metric isn't perfect in several regards, but it is widely comparable across organizations.
Right now, we're stating that we're around 80 to 85% virtualized across our landscape. Some smaller portion of that runs in the IT-as-a-service model, but the gap between "virtualized" and run as-a-service is closing fast.
We can do a good job of quantifying the benefits at each stage of the journey (both "hard" and "soft" metrics), but -- here's the important point -- the ideal measurements change from stage to stage.
In the first phase where you're simply virtualizing non-critical IT-owned applications, it's all about cost optimization: spending less on servers, storage, etc. and the people who have to administer them. Most everyone understands what simple virtualization can do in this regard.
In the second phase, it's more about delivering better IT services under the traditional guise but using virtualization as a foundation. By better IT, I mean "fewer outages" and "better data protection" and "more consistent IT compliance" and "reduced operational effort".
As well as saving some serious money :)
In the third phase, it's really all about being more responsive to the business -- agility -- as well as being able to deliver better IT services (more robust, more secure, more flexible) and, of course, even more cost savings.
One Of The More Interesting Charts
About a month ago, the EMC IT folks started using this chart, showing how -- as agility increased -- the IT spend dramatically shifted from "keeping the lights on" to "investing in new capabilities".
This outcome shouldn't be surprising -- as the back end of IT gets more nimble and efficient at delivering easy-to-consume and flexible services, there's more time, energy and resources in doing the new stuff. And those new and innovative capabilities are always big consumers of the service catalog -- as you'd expect.
As an exercise for the reader, where would you put your current organization on time to provision a developer-ready or user-ready application environment? Not just the server -- everything you need for it to be directly usable for the intended
purpose.
Or, in terms of how much of the IT spend goes to doing new initiatives vs. feeding as-is operations?
Just knowing what the answers might be is a useful exercise.
EMC IT's Cloud 9 Example
Here's just one of the internal services being offered by our internal IT team, dubbed "Cloud 9". Note: a little marketing can go a long way :)
It's self-service infrastructure for anyone who wants it, typically for transient requirements of 90 days or less.
Longer use and/or more demanding requirements initiates a separate process, but if you can stay within the confines of what the service offers, you get what you need -- no questions asked.
The biggest audience? Internal software developers -- both within EMC IT and our various product development communities.
What you can't see from this slide is what's behind it: a small team of people talking to the users of the service, figuring out how they're using it, discovering unmet needs, trying to improve the service, etc.
Put differently, this wasn't your typical hit-and-run IT project; it's part of a sustaining organizational function to deliver internal IT services that people want to consume.
And that makes all the difference in the end.
Being Public About Inhibitors
Change isn't easy, and we're doing everything we can to share not only our internal challenges, but the challenges we hear from our customers who are doing the same. Here's a summarized list.
On the technology front, no real surprises here.
Building and running large, shared infrastructures that deliver a variety of different services is very different than the one-application, one-infrastructure approach we're all so familiar with.
It's not that it's intrinsically hard; it's just different.
As you increase the degree of automation, the people responsible for the outcomes get progressively more nervous. You need to allow time for comfort and experience to develop around a certain degree of automation before progressing to the next level.
Again, it's really human nature complemented by practical experience.
On the operations front, there's a lot to address. You're basically building a new function that perhaps didn't exist before. There are new management processes, new life-cycle processes, and all the skills, roles and org structures that go with it.
The challenge here seems to be directly proportional to the size of the IT operation, with smaller IT organizations having a decided advantage.
The real heavy lifting happens at the interface between IT and the business. There's the need to establish a service-oriented culture when non-IT people interact with IT people. There's the need to teach businesses to consume off the service catalog vs. buying stuff and handing it over to IT to run.
And finance has to get involved to not only support the shared services model, but also take responsibility for managing overall levels of IT spend -- in this model, IT can't be doing that and be expected to be successful.
Advice On Getting Started
Again, every situation is different, but we've seen enough examples where some useful patterns are starting to emerge.
First, there needs to be a clear case made for investing in a transformation.
Going from one style of IT to another isn't simple, easy or cheap. The case needs to be made to executive management, business unit leaders, and the IT organization itself.
More than a few IT leaders have launched themselves on a transformational journey without taking the time for this important step. As a result, they've had to stop, go back, and retrace their path before proceeding again. Or, they've misjudged their readiness to move ahead -- and things end up taking much longer than they should. Either way, a key step in the journey for most.
A second key step is finding business champions outside of IT -- people with a vested stake in consuming the services to be delivered. Not so much as a source of funding, but as your first "target customers" who can validate and consume what you intend to deliver.
A third key step is to break off a small team from the mainstream IT organization, and empower them to move ahead quickly without having to drag the legacy behind them. As they mature their capabilities and processes, they can progressively handle more and more of the IT workloads using the new ITaaS model.
A fourth key step is to design the first round of services for other IT groups who can benefit from attractice and easy to consume services. Almost every time, this turns out to be infrastructure services for application development and testing. This "do it inside of IT first" gets to a quick and visible win without having to change the relationship between IT and its users at the outset.
Finally, be very open and honest about the inhibitors along the way. Change isn't easy, and being absolutely transparent about the challenges ahead makes them a bit less, well, challenging. That's worked for us internally here at EMC; we'd highly recommend it for others as well.
How EMC Can Help
Since we've been working on this for a while, we've amassed a considerable portfolio of expertise, solutions and foundational technologies that can greatly accelerate our customers' journey to a competitive ITaaS model.
We can't cover everything we and our partners do here, so consider this a sampling of highlights.
In addition to the "making the case" and "readiness assessment" consulting work described above, we can dive down deeper into the next layer of tasks. We can help IT organizations construct their first service catalogs, do the consumption modeling, and help get a handle on the financial model as well.
We also have a healthy and growing roster of compatible service providers, which play an important role. Not only can you benefit from understanding from what they're doing (example: learn from their rate cards!), they're an important consumption option that can help accelerate the transition.
More than a few of our customers have started their journey with rented infrastructure and capabilities from a compatible service provider, enabling a quick win to the business while building the case for a more substantial internal investment. They can focus on making the front-end of IT-as-a-service work without having to invest in the back end up front. When it's time, the move is easy. Or the workloads can stay where they are if you're happy :)
When it comes to operations, there's a lot to be done, and we're prepared to help.
We can identify the key skills and roles you're going to need, and various options for sourcing the talent. We can define the operational process you'll need, how you'll get there, and how you'll be continually improving them. We can show you organizational models that work, and help you plan on how to make them work in your own organization.
Finally, there's technology to consider.
From Vblocks to data protection to security to our EMC Proven Solutions; the technology you'll need is there, and it's relatively easy to deploy and operate -- whether you do it yourself, or use our implementation and migration services.
Final Thoughts
Going from a world where you don't have to compete -- to one where you do have to compete -- can be jarring to everyone involved. Whether you're an IT specialist, IT manager or IT leader, it's a completely different game when your capabilities are rigorously stacked up against those of others.
Part of the inherent challenge is simply recognizing what's happening, and starting to do something about it. There are literally hundreds of external IT service providers that compete for enterprise IT business each and every day, and are successful at it. And there's no reason why an internal IT function couldn't do the same -- but only better, due to the natural "home field" advantage.
IT leadership, in particular, is finding themselves especially challenged. The model for IT success is shifting rapidly -- it's now about delivering valuable IT services that the business wants and needs, doing so at a reasonable cost, while maintaining the control that the organization demands.
The good news is that we now have plenty of examples of IT organizations that made the change, and are starting to see the benefits. It's working for them. And those that have made the decision to invest in transformation are justifiably proud of what they and their teams have achieved.
And, make no mistake, we're proud at EMC to be a part of their success.
Who said IT was boring?
By: Chuck Hollis
