State of Change, Chapter 5: Cloud Dynamics and the Structure of Your Company

It’s something I call the nuclear option, and you might not think think I’d lead off with it in an article with “The Structure of Your Company” in its title.  But to deny the presence of this big red button, is to overlook the real reason why the changes in business technology today are so important.  Technology has always been changing, and always rapidly.  But this time, the direction has changed as well.

Cloud service vendors introduce the nuclear option to prospective customers like this:  Your goal is to reduce IT costs.  IT costs skyrocket when you devote a growing percentage of resources and capital toward simply maintaining existing systems.  And there’s no competitive advantage value to be gained from maintenance alone.  Why task your own people with solving the identity and access control problem, and spending your own money to do it, when we’ve already solved it hundreds of times over?  Why task one or two of your IT staff with managing your access control lists a few hours per week, when we have people on staff who do that full-time... and, for that reason alone, know how to do it and are less likely to screw it up?

Or, as more than one vendor has literally put it in these exact words:  “Who needs IT?”

Image taken from a Polish commuter train control room, licensed under Creative Commons 2.0.

Eye of the Beholder

“The core IT departments, particularly in larger enterprises, typically are the guys in the meeting that have to say no to things — who say, ‘That sounds like a fantastic idea, but we need some hundreds of thousands of dollars, and it’ll take us nine months to run through the procurement cycle in order to get the infrastructure that we need to support this fantastic idea.’” remarks Matt Wood, principal data scientist for Amazon Web Services.  He’s talking about the logistics people in an organization whose tasks include explaining to the CxOs why things can’t be done tomorrow.

So when someone new comes along to tell you things can be done tomorrow, it’s hard not to see IT as something of a lead weight.  The new message from Amazon and other cloud leaders is the opposite of the one made three decades ago — that IT is the catalyst for all competitive advantage in an organization.  As Wood explains:

I think the goal of IT, and technology in general, is to act as an accelerant for ideas, for business opportunities, and ultimately for improving the business on behalf of the customer.  And I think in a traditionally, statically provisioned infrastructure, there’s a lot of what you could consider undifferentiated heavy lifting.  There’s a lot of work that goes into building and maintaining and racking and stacking and powering and fixing traditional IT equipment, which does not provide differentiated business benefit.  In the same way as when you are working with electricity, you don’t tend to build and operate your own generators.  They’re expensive and they’re smelly and they reek all the time, but more frequently, if you are a company that exists to serve your customers — a Web startup, or a large enterprise delivering Web-based services to your customers — operating that machinery, that electricity generator is not really your core competency.

Your core competency is in delivering value to your customers.  That might be in delivering more efficient use of pharmaceuticals, developing better drugs.  That may be in building a better user experience for your Web site’s customers, these sorts of things.  So these core competencies don’t translate into the running of electricity generators.

This point of view may appear to contradict, or run completely against, the information technology model of the last decade.  But there are arguments against that statement which, when condensed, sound like this:  The IT department was supposed to have been, ideally, the creator and implementer of new business processes.  But decades of technology buildup has left organizations spending a majority of their IT budgets on maintenance and upkeep, as surveys of businesses of all scales confirm.  So while IT was supposed to have been the fountain of all differentiated business value and competitive advantage, these arguments conclude, that fountain was never turned on.

Peter Coffee, a former press colleague and now Vice President of Platform Research for Salesforce, explained it like this for a 2010 video for his company’s blog:

Cloud means two things:  It means that capital budgeting is no longer part of IT provisioning.  It means you do not have to begin the conversation about meeting a business need, solving an enterprise problem, entering a new market... with, “Well, what is in the capital budget?  Maybe we can do this next quarter; maybe we can do this next year.”  As far as I’m concerned, it is essential to the label of “the cloud” that if you know what you want to do, you can begin to do it right away.

The other aspect of the cloud that I think is absolutely inseparable from the label is that, if you are licensing and installing and patching and updating a big, hairy, complicated stack of software that is in no way creating competitive advantage for your business, but is merely the stack of commodity stuff that you have to keep running before you can add business value on top of it, you know what, if you’re doing that yourself, you are not practicing cloud computing.  You are practicing software infrastructure maintenance.  It’s expensive and time-consuming, and it does not create competitive advantage.

The SISP revolutions were attempted on the basis that innovation was something that could improve business processes, among the most important of these being maintenance.  Organizations could conceivably discover business value through improving the methodologies they used to keep their core processes efficient.  Practitioners of the new revolution postulate that every minute you spend performing maintenance is wasted energy, and that business value only derives from doing something else.  Thus, theoretically, there is no innovative maintenance process from which business value can be derived.  The only way to generate business value, therefore, would be by outsourcing all of it.

In their 2008 book The New Age of Innovation: Driving Co-Created Value Through Global Networks, C. K. Prahalad and M. S. Krishnan attempt to sew the seed of the new IT revolution by coupling Porter’s original notion of business value with the whole notion of the “eye of the beholder.”  Their “brand” is the kind that a rancher makes on a steer: a simple pair of formulas that are easy to remember and print on a T-shirt.

Their beholder is the customer who, in their model, is the sole person responsible for judging business value.  The formula N = 1, therefore, stands for the number of customers every organization should perceive itself having.  Value can only accrue, the authors argue, when each customer perceives herself having a unique, personalized experience — rather than sharing something with the world, having something no one else in the world has (but bragging about it to the world on social media later).  The other formula is R = G, which is shorthand for “all resources are global.”  Their argument proceeds that, since it’s the customer who determines value in the end, whether your organization owns the resources it uses to deliver that value is inconsequential.

Put another way, ask your customer what she thinks of your IT department.

Amylin Pharmaceuticals was one of the earliest adopters of cloud and grid technologies, dating back to 2000.  Speaking on a panel at a recent CommNexus conference in San Diego, Amylin CIO Steven Phillpott told attendees that Prahalad’s and Krishnan’s book changed his company’s perspective.

Phillpott gently brought up the subject of repurposing internal IT, relieving it from the day-to-day affairs of those tasks his company nicknamed “KTLO” — “Keep The Lights On.”  He explained that the R = G formula enlightened his company to the realization that it will never own all the resources it needs to provide personalized pharmaceutical experiences to his customer.  “You are going to have to go to external resources,” he said, “in order to leverage and support your customers.”

It follows, Phillpott explained, that a company only needs to access resources, not own them.  “I think once we had that concept in our mind,” he continued, “it allowed us to adopt the cloud methodology, the cloud paradigm, a lot easier than a lot of other organizations that still struggle with, ‘Well, how do I define cloud?’  We didn’t care how you define cloud.  We were like, ‘How are we going to use a service, whether we own that service or not?’”

“Businesses of any scale. . . will, at some point, discover that it is impossible to totally outsource a fundamental function of the corporation, especially when the distinctiveness of that function is critical to its attainment of competitive advantage.”

So the Amylin strategy was to outsource its entire data production facility to a remote facility, to lease infrastructure from Amazon and applications from Salesforce’s Force.com and Heroku platforms, and to lease storage from Nirvanix.  Amylin’s IT workforce could then concentrate, he explained, on the tremendous problem of how to move applications to the cloud while concentrating on government mandates for compliance.  This way, IT could concentrate on four areas that could be considered “pillars:” data, process, reporting, and analytics, which Phillpott says are the areas where IT delivers value to the business.

“The people who are running data centers are doing it so much better than we could ever imagine,” says Phillpott.  “Why do we even want to compete with that?  E-mail, HRS applications, CRM applications.  Hey, let the experts work on those, and we’ll focus on the things that are important... things related to drugs and drug discovery.”

Though Amylin was acquired by Bristol-Myers-Squibb in August 2012, the company and its new parent remain among Amazon’s star customers.  Amazon’s Matt Wood tells me about clinical trial simulations that BMS has run on AWS infrastructure.  One pediatric study, for instance, can be run using fewer patients spending less time in hospital.  At the same time, the total compute time consumed comes in at two percent of the time the same simulations consumed on-premise.  As Wood explains, with experiments at this level being more prone to human error than other types of tasks, they incur more direct and more painful costs for organizations forced to run them entirely on-premise:

Fundamentally, you’re going from an artificial constraint of information technology that requires this large capital investment, to a removal of that constraint.  And that enables your existing staff to go off and deliver the opportunity of IT, the promise of IT, to accelerate your business, rather than have this artificial constraint in front of them.

It’s being able to respond more quickly to your customers’ needs, to use the information in the data that’s being generated through your customers interacting with your organization, to be able to quickly respond to their changing needs [and] to the changing environment of your organization.  IT is enabling all of that, because we’ve started to remove some of these artificial constraints around infrastructure.

Does IT Have Value?

The plan sounds viable:  Outsource all the KTLO tasks to the cloud, and repurpose IT to become strategic once again.  The leadership role would seem like what Synnott and Gruber had intended back in the late ‘70s with their original concept of the CIO: a kind of “race strategist,” someone who determines the best balance between control and growth.

One of Salesforce’s cloud platform components is Chatter, which tech journalists often confuse for a kind of “Skype for business.”  It’s actually both the collaboration layer between platform users and the inter-application communication layer that substitutes for middleware for Salesforce apps.  In Salesforce’s vision of the redefined organization, Chatter provides a virtual workspace for employees at different levels, conceivably obliterating the compartmental barriers between business domains.  But also, it conceivably makes the CIO into the business counterpart of the social space’s community leader or group moderator.

“The really fun stuff with technology isn’t maintaining infrastructure,” remarks Anna Rosenman, Salesforce’s senior product manager in charge of Chatter.  “The fun stuff is building tools, building applications on top of the social and mobile platform that elevates your business, that empowers your co-workers and yourself.”

I asked Rosenman what she expects should happen to an organization that outsources all of the operational jobs that consume three-fourths or more of IT’s time, to a CSP.  Can an organization just switch IT’s tracks this quickly without repercussions on the company as a whole?  What’s to keep a company’s executives from taking the nuclear option?

What I think the cloud actually does is create a better relationship between IT and business.  Should we get rid of our IT organization altogether?  Frankly, no.  IT professionals are critical to the success of a company, so we can help IT professionals do the work that they prefer doing — educated people with a wealth of experience who want to be there to make that company successful.  Strategic IT doesn’t come from building infrastructure.  It comes from building applications, building technologies, connecting systems together in such a way that they empower business needs.

This is where the present splits from the past:  Prior to the dawn of SISP, “I/S” was considered the fundamental infrastructure of the organization, and “I/T” the tools built upon that infrastructure.  SISP was instrumental in melding the two together — in declaring that all strategy was fundamental, and therefore all strategy was infrastructural.  The goal was to enable a strategic element of the organization to take charge of the undercurrents, to make sure the operational elements always have the resources they need to execute principal business functions.  It’s from the successful execution of those functions, SISP declared, that businesses derive competitive advantage.

Cloud vendors in the modern era have painted over this message entirely.  Infrastructure, they say, is no fun, a waste of time, and most critically of all, unimportant to how your customer perceives value.  As long as platforms are no longer distinguishable from one another, logically, you must conclude no competitive advantage can be derived from them.  Thus the traditional concept of IT goes away.  Salesforce and others prefer not to characterize the act like the cutting off of one’s own arm.  But in a world where N = 1 and R = G, IT = 0.

IBM is one of those vendors I talked about earlier in this series, with an interest in sustaining its brands and therefore not so much of an interest in disrupting the order of things.  The people in an organization most responsible for sustaining the value of IBM are those in IT who have depended on IBM to provide infrastructure.  Over the past few years, IBM has advocated the creation of a new CxO role, which it has often referred to as “Chief Data Officer.”  In companies that already have CIOs, how could they co-exist with CDOs?  Or does the CDO replace the CIO?

I asked Ric Telford, IBM’s vice president of cloud services, won’t companies always need someone in charge of infrastructure?  Though he answered yes, Telford said this person may not necessarily have to be a company executive, especially for smaller firms — perhaps a consultant could fulfill that role.  However:

I think in any company of size, there’s always going to be a need for a CIO organization.  In a small company, if you say they still need the equivalent of a CIO, even if the CIO is someone who just happens to be the most technically [skilled]... Somebody has to make the decision they’re going to use IBM SmartCloud e-mail and collaboration versus Microsoft.  Somebody makes those decisions.  Sometimes in a company, it’s just a financial and procurement decision; other times, it’s part of a strategy.  I’d like to think that any company of size is managing it as a strategy, and they have some semblance of a CIO organization or individual who is doing these things with a strategy in mind and with purpose, rather than just going with the first person who knocks on the door who says they can do e-mail in the cloud.  There will always be somebody who rises to the top, but in any company of size, that’s what we call the CIO.

But it’s more of a strategy than operations.  That’s what’s happening:  A lot of the operations side that consumes internal IT today goes away, or goes to a public cloud service provider, allowing the people who are already experts in the company to think more strategically.

Note Telford’s use of the phrase “CIO organization” as opposed to an individual — suggesting that any number of people, perhaps more or less than 1, could fulfill the role depending on the company.  Theoretically, a few of the people whose jobs Telford perceives as being consumed by the cloud, could move into this “CIO organization.”

Nonetheless, does the trend of “IT consumption” spell trouble for IT professionals whose job skills and certification are based around operations rather than strategy?  Maybe infrastructure is no fun, as Salesforce puts it, but some people learned to do the heavy lifting because they wanted the big bucks.  If there’s no heavy lifting to be done internally any more, do we really expect any organization of scale to graciously re-educate its heavy lifters?  IBM’s Telford tells me he frequently meets people who were hired to do strategic planning, but end up spending their time with re-syncs, reboots, post-mortems, and outages.  “Those are the people who can really benefit,” he says, “because they can get back to what they thought they got hired to do, which is make IT a business differentiator for a company.”

Yet this job reassignment, Telford concedes, is mainly confined to what he calls “enterprises of scale.”  For smaller organizations, he admits rather cautiously, the availability and reliability of infrastructure from public providers will make redundant any level of on-site IT operations leadership.

I do believe that, in the small and medium business cases, there is going to be a point where they just get out of the IT business altogether, and everything should go to the public cloud.  The idea of the nurse or the receptionist in the doctor’s office whose responsibility includes the rebooting of certain servers, and backing up e-mail systems, I think will be a thing of the past.  I just don’t believe there’s a compelling value proposition for servers to be co-located in retail stores or doctors’ offices or lawyers’ offices, small commerce businesses, in the future.  With high-speed bandwidth and cloud computing, and public cloud service providers, I think small businesses are going to be totally relying on public cloud in the future, and enterprises of any size will have hybrid IT environments for as far as we can predict.

Not with a Bang

Businesses of any scale are reconsidering whether they need an IT department at all.  While it sounds magnanimous to suggest that IT personnel will be “freed” to devote their energies to “the things the company does best,” if we’re being honest with ourselves, the only reason any organization considers the benefits of eliminating IT is to exploit the benefits of trimming down the payroll.  For some organizations, such trimming may be necessary for their survival.

But all of them will, at some point, discover that it is impossible to totally outsource a fundamental function of the corporation, especially when the distinctiveness of that function is critical to its attainment of competitive advantage.

Here is the unforeseen, even unacknowledged, problem — the factor that could yet derail the cloud revolution:  No matter how, or to what extent, corporations are disrupted, there are certain “laws of physics,” if you will, that do not change.  One is that business functions are organized by their sources of funds.  While it sounds inspirational to potentially reinvigorate IT by relieving it of operational responsibility and making its job entirely strategic, the whole reason for considering cloud in the first place is to shift its funding source from capital to operational expenditures.  If, as Ric Telford states, there will always be someone in an organization making purchasing decisions.  If those decisions essentially consist of which brand to go with at what time, then although they may take strategy into account, they are fundamentally operational decisions.

The person making those decisions is the opposite of the CIO.  You don’t convert a strategic function into an operational function without subsequently substituting the strategic framework with an operational framework — without replacing the 18-month vision and the three- and five-year vision with the 30-day and 90-day vision.  This narrower term may indeed be the more appropriate term for cloud services, and may make these services not only more efficient but far easier to manage.  But it creates a disconnect between the strategic centers of the company where competitive advantage is conceived, and the operational centers where that advantage is generated.  And that same disconnect has already been blamed for the failures of information revolutions throughout the past.

If IT is, as the SISP leaders suggested decades ago, not only a fundamental business function but the principal generator of competitive advantage and genuine value in their designated markets, then how can the elimination of that function be any less than a public self-beheading operation?  In those many cases where SISP did fail, it was not because the observations of its creators were wrong.  It failed in organizations where the motivation for change was never strong enough to attain critical mass, to launch the revolution.

SISP’s problem was the lack of alignment between IT and management.  Cloud opts to resolve that problem with the nuclear option: the complete outsourcing of the IT job, and the elimination of that department.  Nothing left to have to align.  Cloud service providers are telling customers that cloud dynamics moves their competitive advantage generator off-premise.  Therefore, the competitive value of having on-premise staff maintaining policies and procedures, becomes zero.

Obviously, no one wants to spend operational funds, or any other funds, on a function that produces no competitive advantage.  While that makes sense, it must also follow that competitive advantage cannot manifest itself from a business function allowed to run on automatic pilot.  Truly effective cloud service providers do not frame their services in such a light, but rather as enablers of IT in both the strategic and operational senses: as balances of form and function, the way the creators of the CIO role originally conceived it.  But many CSPs are enticing customers with a vision of nothing left to have to balance, framing cloud as a panacea.

The first IT revolution began with the notion that a company’s information technology department may be the very fountain of its competitive value.  The cloud revolution is being led by advocates who declare that fountain has run dry.  This is the quintessential difference between this movement and all that came before it.  This is what assures us that the aftermath, one way or the other, will be measured on a colossal scale.