I consider it an article of faith that many enterprise IT organizations will need to learn to compete for their internal customers.
If you've been following this blog, you know I've been drawing heavily on the experiences of EMC's own IT organization as they progress through this inevitable transformation.
Today represents an interesting milestone of sorts: the formal internal launch of a "critical mass" of self-provisioned IT services that anyone at EMC can freely consume.
Yes, they've had many of these internal services out there for a while, but now our EMC IT team has started to actively encourage consumption at scale.
I think this particular event serves as an excellent example of how the thinking changes when enterprise IT goes from being seen as a monopoly to being seen as an aggressive competitor.
If you work in an IT setting that's starting to go through some of these changes, I think you'll find the discussion here very enlightening.
Our Story So Far
Our EMC IT team is thankfully somewhat ahead-of-the-curve in its transformation to becoming an internal IT service provider.
Most of the back story can be found in a popular series of posts ("Leading An IT Transformation", "Organizing For Success", "Achieving Financial Transparency", "Cloud And The Act Of Being Selfish", "Why IT Groups Will Invest In Marketing" and "Price Signals And Why They Matter".)
The key motion is the formation of new IT functions that create, introduce and manage a portfolio of easy-to-consume variable IT services that are attractive and competitive.
We've got an important advantage that many IT shops don't have: the back end of EMC IT's "service factory" is very standardized, e.g. built on Vblocks. That decision was made a long time ago :) This standardization has enabled our IT team to spend more time on making IT easier to consume vs. spending endless amounts of effort on the back-end plumbing.
Although the majority of IT services (e.g. infrastructure, platform, etc.) are consumed by other IT groups in support of project-oriented activities, there's been a progressive exposing of IT services directly to end users across the EMC organization.
During 2011, there's been a lot of "trialing" of directly-consumed IT services -- pushing out various services directly to end users to gain experience in relatively new disciplines: pricing, marketing, self-service consumption, self-help support, back-end operations and the like.
Crossing The Line
Like any large-scale product launch, there comes a time when you say "we're ready for the big time", and you focus your efforts on driving awareness and consumption vs. getting more comfortable with what you're doing.
This morning, I came to work and saw clear evidence of a formal "services launch" by EMC IT to its 50,000+ user internal community.
For all intents, it looks very similar to how an external IT service provider would "launch" to its own customer base.
And I bet we see far more of this from enterprise IT organizations in 2012.
What Was Launched?
Although EMC IT has been doing various cloud-like IT services for a while (e.g. lab as a service for our internal developers, desktop as a service for those trialing that service, etc.) this morning's launch was different in several regards.
- The services are now launched to each and every EMC employee, regardless of organization, geography or role.
- Each of the services had been technically available for a while, and each are based on proven and mature processes
- Clear expectations have been set around target use cases, pricing, competitiveness, chargeback and service level.
- EMC IT felt good about being able to meet the demand side (both process maturity and resource availability), so this was a distinct move to increase and drive consumption.
I am willing to call a trend, and say that "aggressive launch to drive demand" is one of the key signposts along the journey to IT transformation.
It basically says you've arrived: your focus is now shifted to aggressively competing for the business.
What IT Services Were Launched?
First, there was a more formal launch of Cloud 9, a self-service IaaS offer for non-persistent workloads. The idea is simple: anyone in the EMC organization can spin up a decent amount of infrastructure and various software stacks using a self-service portal. No questions asked.
It's 100% free, but it's only good for 30 days.
The motivation was simple: we had a lot of people going outside the company for easy-to-consume infrastructure services. Besides the obvious security concerns, there was a nasty problem involved when workloads became more mature and had to transition to EMC's internal IT environment.
Now, there's no reason whatsover to go outside. During early trials, over 1000 workloads were created on Cloud 9. That's over 1000 workloads that didn't go to $deity-knows-where -- and that was just during the trials.
The second announced service is a persistent IaaS offering (marketed under the catchy name of "IaaS") that is targeted at non-mission-critical workloads. The internal service is clearly marketed as "expect the possibility of an outage lasting as long as 24 hours". If that doesn't work for you, don't use this service :)
Four things of note here.
First, there's a rate card right up front that is very clear as to actual charges: what's included and what's not. For example, since you're using the internal EMC network, there are no variable charges associated with data transfer. Don't know if that can last forever, but that's what it is today.
Second, there are several pricing examples showing cost-to-service on the EMC IT IaaS offering vs. competitive alternatives such as Amazon, AT& and Rackspace. Let's just say that the internal pricing is very competitive -- so much so that no one could make a reasonable case that it's "cheaper" to go outside.
Third, the TOS (terms of service) are there for all to see: mostly written in plain and easy-to-understand language vs. legalese. After all, it's one part of EMC providing services to another part of EMC, so we really don't need to get the lawyers involved :)
Finally, there's a rather thoughtful explanation around chargeback: what it is, how it works, and why it's good for EMC IT consumers.
In previous IT models here at EMC, there was only a casual connection between what you consumed and what you might pay; in this model, you get a detailed monthly invoice tied to separate cost centers created for the purpose.
This is essentially a new IT consumption behavior for many here at EMC; so it makes sense to overinvest in user education around the topic.
The other service announced was DocBox, a self-service enterprise-grade collaboration and content management environment based on Documentum, modeled somewhat around Dropbox and Box.net. Since it's based on IIG products, it provides a richer set of functionality than you'd expect, but that isn't really reflected in the pricing.
The same four elements found in the IaaS launch materials are found here. The pricing examples are illustrative: richer functionality doesn't cost any more than popular external alternatives. And EMC gets the benefit of securing its own information :)
There's also an interesting preview of BI (business intelligence) as a service. Power users will be able to easily spin up their own analytics environment using the same model as described above. During the first part of 2012, we'll also see UEaaS (user experience) as well as a few other cool offerings added to the offering.
How Do You Know These Are The "Right" Services?
The trick isn't the actual services in the portfolio; it's the organizational functions that led to their creation. Our internal ITaaS team is talking to internal users, figuring out what they want, creating services they want to consume, measuring their uptake, and re-formulating the portfolio over time.
Simply put, they're measured and incentivized to "get the business".
If these particular offerings aren't the right services, new variations will quickly be created: using the same processes and plumbing that created the current ones. Just like any competitive external IT service provider, you've got to be good at giving users what they want -- otherwise someone else will.
Now that the basic operational processes are in place (and proven!), I'm guessing we'll see a fast-and-furious pace of new IT service introduction and modification during the first part of 2012.
What If I Need Something That Isn't On The Menu?
The traditional project-oriented interface to EMC IT hasn't gone away -- if anything, it's stronger than before.
First, business-facing IT representatives now have a decent services catalog that IT users can consume directly for a decent part of the new projects coming down the pike. That means that the business users can often get started immediately on prototyping and assembling the pieces they need (using the standard service catalog) without the need to define and justify a substantial IT project.
Everyone wins as a result.
Second, when a familiar IT project rears its head, it's supported by the same sort of easy-to-consume variable IT services that are provided elsewhere. Same factory, different consumption models. The preliminary stats on the resulting project acceleration (old vs. new model) are quite compelling in themselves.
Third, since all IT services are provided off essentially the same infrastructure using the same processes: there's amazing scale and standardization which, of course, results in lower unit costs for IT and ultimately better services delivered. At EMC, IT can often compete solely on price if it comes to that, simply because they use the same techniques that service providers do.
Also, let's not forget that our internal IT guys have two key advantages that external service providers don't have. First, there's no need to make a healthy profit. And, second, the EMC IT team can make the justifiable claim that they know EMC's business than any external service provider could ever hope to :)
What If Internal IT Can't Do What I Need?
Another interesting evolution in our internal IT function is clear adoption of the "builder-broker" mentality. EMC IT realizes they can't be all things for all people across a diverse and thriving global technology company. But what they *can* do is intelligently broker services for the pieces that have clear justifications for going to external providers.
Interesting to note, though -- that list of "gotta go outside" projects appears to be way, way down over the last year or so :)
How Are You Teaching IT Consumers To Pay For IT?
About half of IT services that EMC IT delivers today are charged back in some form; and about half aren't. Even for those that are being charged back, it's a rather obtuse "big chunk of IT" model vs. the more granular and variable types of service consumption, where decisions and approvals are made at lower levels in the organization.
I think the framing is appropriate: essentially, it's a new skill that has to be learned by IT consumers. They need to learn to understand how their requirements might change over time, what they need at the outset vs. what they might need later, and -- in general -- get comfortable consuming off of a service catalog vs. the traditional project-oriented funding model.
Like any new skill, there are populations that get it quickly, those that take a little longer to get comfortable, and a significant population of folks who will inevitably struggle. It's good to see that our IT leadership understands the nature of the challenge, and is prepared to invest over the long haul to fundamentally change IT consumption habits across the company.
But -- What If ... And What About ... ??
The reaction is sort of predictable. I start sharing with an IT leadership team how our own EMC IT organization is well along the way in its transformation, and starting to look more like an internal service provider and less like a traditional enterprise IT organization.
I've learned to shut up after 30 minutes or so -- just to let all the anxieties bubble up.
What if there's more demand than you have supply? What if your users don't use your new service? What if the finance guys don't go along? What happens when the service isn't available? Etc. etc. etc. It's quite a long list of potential anxiety bubbles indeed.
I spend a lot of time patiently responding to each and every concern: here's how we thought about that, here's what we did, here's what we learn.
But I only do that for a short time.
At some point, I draw a line and attempt to simplify the discussion.
Our EMC IT team decided to change the way they did business.
Have you?
Excellent post!
Posted by: Shahin Khan | December 23, 2011 at 06:44 PM