Where's the Business?

Despite all my attempts it appears cloud computing continues to be a technology topic, at least in the Fortune 1000. Over the past few years secrets of cloud, including Facebook's private cloud, Zynga's use of 12,500 Amazon EC2 instances, and NetFlix creation of their Chaos Monkey, have made big company executives ask their CIO "what are we doing with public cloud? What are we doing that we can tout in Fortune magazine or at our next board meeting. After 30+ years of being beaten into submission that IT is about avoiding risk, now CIO's are being asked how far outside the box they're thinking, how entrepreneurial they are.  I'd be amazed if none have broken down in heaving sobs or curled up on the floor with their thumb in their mouth.  The truth is, cloud doesn't start with the CIO, it ends at the CIO.

The preconceived notion is that cloud is all about TECHNOLOGY. Guess what, it's not! In reality each of those examples, and the ones CEO's and CFO's trot out to "motivate" their CIO's have nothing to do with technology. Rather, each has to do with the business. It's the needs of the business which drove the application of the technology, not vice versa. Therefore the right question is not CEO to CIO but CIO to the CEO: "Here is what cloud enables businesses to do differently. How can we take advantage of it?"

As clouds are being built, the enemy of efficiency, fiefdom, is following right along. With the number of business savvy CIO's in the Fortune 1000 I'm surprised at how few have engaged the business in a discussion outside of cost savings. Great, run applications and all at a lower cost. But that wasn't the goal of Facebook, Zynga, Twitter, or just about anyone other cloud based company.  Just look at NetFlix who single-handedly put Blockbuster out of business and forced the cable operators to take notice.

Cloud is a platform, a new business enabler for the 21st century.  It's about revenue generation,  It's about providing an on demand foundation to deliver technology where needed, when needed, in the most efficient manner possible.  How it happens isn't important.  The focus needs to be on WHY!

(I held on to this post for a year or two before publishing to validate it's applicability.  Then, all of a sudden, I started seeing a torrent of posts and articles on this very subject.  So then I held on to it for a few more months.  But finally I've come to my senses, so my apologies for helping to push the conversation earlier.)

Where is the Hybrid Cloud Adoption?

Since the early days of the enterprise cloud adoption discussion the concept of Hybrid Cloud, combining the control and security of the private cloud with the breadth and economics of the public cloud, has been a topic.  Cloudbursting emerged as one of the early, highly touted use cases for cloud, and I was one of those consultants adding to the hype.  Yet today with valid reasons for adopting Hybrid Cloud, where are the adopters?

I believe there are several reasons Hybrid Cloud has yet to take off including:

• Offerings - the offerings in the market fall well short of what companies need because none provide any orchestration, policy management or governance tools.  Buyers need a way to specify the environment blueprint to automate the provisioning at the provider.  The environment needs to be  configured precisely as needed in an automated fashion and this must include the ability to specify the location based on the location of other virtual machines and storage.  Policy management across providers, applications and data are necessary to ensure data is protected, business rules are enforced, and costs are effectively managed.  And without governance tools in place administration and change management become ad-hoc to a large degree.  There are tools available to help, but without strong provider integration the tools are merely best efforts.
• Critical Mass - there are companies who are using Hybrid Cloud successfully today, but they rarely talk about it.  Whether its to maintain a low profile to the public and shareholders in case the bet doesn't pay off, to protect competitive advantage, or simply because they don't feel others should benefit from their investment we have very few stories to learn from and retell.  Most of IT progress is predicated on following the leaders.  CRM, SFA, ERP, MDM, eCommerce, Web 2.0, Mobility, M2M, Analytics.  All of these technologies are pioneered by visionaries and derided by the status quo until a tipping point is reached after which the critical mass of thinking drives widespread adoption.  However in Hybrid Cloud the benefits of sharing the stories in the form of larger, more robust Public Clouds at lower costs points will benefit the early adopters.  Telling the stories will yield dividends.  There is very little competitive threat of the followers suddenly catching on and accelerating into a pioneer.
• Risk vs Reward - we know many CIO's have no interest in doing cloud let alone leveraging public clouds because they don't see the reward for taking the risk. Risk is nothing more than a euphemism for fear: fear of failure, fear of losing control, and fear of personal risk. Although all the experts say it's the only way to go and a fundamental part of transforming IT into a business services organization, cloud isn't easy.  And anything that isn't easy is rife with risk.  Therefore it remains easier to follow along at a snails pace talking up the desire to leverage Public Cloud resources in a hybrid model than actually take any steps.  Often these organizations are characterized by having a less than robust private cloud and have done no organized research into making a Hybrid Cloud reality. 
• Learning Curve - learning anything new takes time, and with an expansive domain such as cloud becoming an expert seems impossible.  Focusing on private cloud enables organizations to build skills related to cloud and move up the learning curve at reduced risk because there is no threat of losing control.  However Private Clouds provide a false sense of accomplishment because the end of their road is rarely much further than mass virtualization, well short of the value propositions of cloud.
• Business Value - the concept of cheaper, better, faster has made inroads at the virtual machine level, but few in the business understand how Hybrid Cloud can drive real business value and thus are allowing IT more leeway than necessary.  The last thing many IT leaders want is a business to expect the deployment of entire environments within minutes, scale up and scale down within minutes, and frankly deployment of entire new capabilities in minutes.  It makes IT look slow and incapable.  Slowly but surely IT executives are getting their arms wrapped around the idea that Hybrid Cloud is an enabler, but at the same time they have to gain comfort with giving up control.  

Just as cloud isn't easy, neither is transforming IT or being an IT executive at this point in time.  Control has been the hallmark risk mitigation technique of IT for 40+ years and is unlikely to change soon.  And why is this the case?  Because the business wanted it that way.  Consider it the unintended consequence.

The Blinding Light of Big Data

I have held back from publishing this article because I'm not sure I'm using the right terminology.  I am not a data wonk.  However over the past six months my interactions with those who are heavy weight experts have convinced me I'm on the right path and need to start the dialog.  So my apologies if my point isn't crisp and please ask me to explain what's not clear.

I feel the bright light of interest in Big Data is casting a long shadow over the convergence of two significant challenges: depth of executive experience with analytics and their understanding of the limits on predicting the future.  Good executives want data to drive their decisions and big data promises to expand the range of leaders using data to draw conclusions.  However I see nothing being done to protect us from a group of leaders thrust into a dependence on data analytics without insight into how to compose analytic groups, an appreciation for the limits of predictive models, and knowledge of the law of unintended consequences.

One glaring reality today is the lack of available analytics talent; we simply don't have enough people with the requisite skills.  The subset of people with the requisite background and experience who can turn the torrent of new data into value is significantly smaller than the need.  Equally important, that subset is not exclusively statisticians and modelers.  We need diverse teams to protect us from weighting conclusions too heavily based on natural biases of one group of professionals.  Picking on one of those groups of people, my favorite quote  "A statistician is a person who can draw a straight line from an unwarranted assumption to a forgone conclusion" reminds me every professional has their own set of biases.  A diverse steam will improve the quality of data used, models developed and conclusions drawn.

Second, leaders need to understand the limits of predicting the future.  We often find data which purports to predict the future through the application of some algorithm which was likely developed through the use of historical data.  But how well does the algorithm's predictions predict the future?  It depends on how much the future repeats the past.  As long as things stay within tight bounds, the future is reasonably predictable.  Yet few leaders understand the stress points of predictive models, the data elements where small changes can generate a wide variance in results.  Our financial meltdown in 2008 was precipitated by an untested boundary condition; what happens to the value of Mortgage Backed Securities (MBS) when the default rate goes outside of historical trends.  Unfortunately we all experienced the result.  Whereas in statistics we were taught such testing is imperative, for some reason it was entirely ignored.

Predictability is the byproduct of repetition without variance.  However today we live in a world rife with innovation.  When innovation happens, analytics are of little help to comprehend the impact until the population size is significant enough, by which time the opportunity door is closing fast.  Once a model seems to predict the future it seems all consideration for its continued accuracy evaporates; worse when it directly correlates to making money.

Finally we cannot overlook the impact of unintended consequences.  There is a well known story of a retailer using analytics on purchases to identify a woman was pregnant, only to inadvertently notify her family of her secret by sending her targeted marketing products.  This is one of many unintended consequences of making decisions purely based on data.  I am convinced concern about unintended consequences needs to be part of the thinking.  As William Gibson said "the future is already here but unevenly distributed".  Having a diverse team is a key element of limiting this negative side of analytics, however executives need to be sensitive to it as well.  One angle I use is to consider the motivation of participants versus the actions being considered based on the conclusion.  If the motives don't match the actions we open the door for an unintended consequence.  Returning to the retail example, the action to communicate marketing information did not align with the motivation of the buyer; the conclusion that all buyers want to save money was false.  Where the retailer failed was in not considering the buyer's motivation which was not possible given their limited data set.

My greatest fear is the widespread movement evolving which attempts to replace intuition with modelling.  I've already seen the seeds of this being sowed in books and articles from highly respected authors casting doubt on the value of intuition.  Intuition helps one to see the future whereas data can only help one to project the past into the future.  The difference is significant.

L-Earning Your Way Up the Cloud Learning Curve

Invariably when asked what lessons I can share in cloud adoption I refer to Jurassic Park.  During the initial tour of the facility, Dr Ian Malcom (Jeff Goldblum) sounds a warning to Jurassic Park entrepreneur John Hammond (Richard Attenborough):

“You read what others had done and took the next step.  You didn’t earn the knowledge for yourselves”

Having spent the majority of my career in emerging technology I have found one immutable truth: you can’t leap and hop your way up the learning curve, you have to earn the knowledge for yourself.  The knowledge you earn provides the critical insights necessary to make course corrections and deliver a solution not only relevant to the existing problem, but one which becomes the foundation for the next solution.

Big public cloud providers like Amazon, Google, and Microsoft earn their knowledge every day delivering solutions to buyers in IT and the business.  In contrast most Fortune 500 companies have chosen to let others earn the knowledge by partnering with private cloud solution providers, who themselves imitate public cloud providers.   Overlooked in the analysis of public or private are the combined powers of expectation and momentum.  What the public cloud delivers today quickly becomes an expectation of the private cloud tomorrow, and momentum is widening the value gap as public cloud product cycles continually contract.  As a result private cloud solution providers are not only well behind on the learning curve, they are falling further and further behind.  This puts IT leaders, who are following the followers, in a difficult position where their only choice is to try and leap up the learning curve on their own.  The results are predictable and all too real: projects fail to meet expectations.  And as the gap grows, pent up demand is beginning to fracture the status-quo dam carefully constructed by IT leaders over the past 40 years.

So what to do?  Realizing CIO's are largely out-manned and out-gunned I figured it's time to put-up or shut-up.  So after years of referring to it in conversations and presentations, I have finally documented “Brian’s Non-Scientific But Anecdotally Accurate Cloud Learning Curve”.  These are the "Ah ha!" moments I've documented in my 10+ year journey through distributed computing in the Fortune 500, from P2P and grid to cloud.

As companies move their way up the knowledge curve they invariably make the same mistakes at the same points on their journey.  The first common mistake is not recognizing cloud is much more than just infrastructure.  Although the ability to provision servers, storage, and networks is important; it's just the foundation.  This leads to the second most common mistake which is not getting the development side of IT involved BEFORE building the "cloud".  Most private clouds at large organizations end up being little more than mass virutalization projects, unable to lead the divide to cloud because of an infrastructure centric approach.  Everything about cloud should be about the application.  After all, why does the server, storage, or network exist in the first place?  Finally the third most common mistake is companies biting off more than they can chew and trying to leap and jump their way up the curve.  Many of those infrastructure centric clouds have, in fact, tried to leap over native application development and enterprise PaaS to deliver orchestration and provisioning and brokerage.  However because the needs of the applications are ignored the scope is too narrow and tools are selected which may not meet future needs.  It's quite easy to paint oneself into a corner when painting for the first time.

Cloud Computing has rapidly emerged as the foundation of the Software Defined Enterprise.  Getting to cloud first and fast sounds compelling.  In reality the path to cloud nirvana is littered by the hollowed out shells of failed projects, primarily due to companies trying to jump the learning curve like John Hammond and derive the benefit without earning the knowledge. 

SDN and NFV are Crucial to Getting Cloud Right

Go back to 2008 and you'll find the beginnings of an incredible cost savings opportunity in the emerging world of cloud computing.  Built on the back of grid computing, cloud offered to reduce operational costs by 30% just through the application of mass virtualization technologies (virtual machines, server and data center consolidation, etc.).  Over time the return on investment improved, until in 2012 executives started talking about a 70% overall cost reduction versus previous generation approaches.  By contrast SDN became a topic in enterprises in 2012 and yet within one year executives at some of America's largest companies are touting the opportunity to reduce network expenses by 70%.  It's no coincidence.

In the emerging SDN/NFV discussion the eye-opening "Ah ha" moment repeating itself almost daily is the realization that when moving network functions to software the platform answer is cloud.  As I referred to in my August 2010 post Everything Should be Software, the opportunity is tremendous to continue the commoditization of hardware by moving specialization into software.  Cloud is the platform for SDN and NFV which in turn enhances the value propositions of cloud:

• Improved agility to meet emerging business demands
• Reduced time to market
• Greater efficiency
• On-demand scalability
• Variable cost

As someone who has advocated for years that cloud is the platform of the future onto which companies will build their eCommerce, Big Data, Mobility and Social/Collaboration solutions; an improved value proposition makes cloud all that more enticing.

Imagine being able to bring the transformation, normalization and aggregation tasks from analytics down into the network.  Establishing secure network tunnels on the fly across wired and wireline networks to connect mobile applications to data sources without the risk of interception by a third party.  SDN and NFV bring the functionality required to realize the new data architecture I advocate.

Of course none of the promise of SDN and NFV comes free.  Limiting the speed of adoption is the same culprit that's been holding cloud back for the past 5+ years: expertise.  We simply don't have enough networking experts with backgrounds in software, just as we don't have enough operations, administration, server architecture, storage architecture and security experts with software backgrounds (equally we have few software people with strong network, infrastructure, and security backgrounds).  We are still hampered by the breadth and depth of knowledge required to get cloud right.  Add to this challenge the cost and complexity of moving networks with 99.999% availability and higher into a new technology architecture and the reality sets in the effort will be herculean.  Luckily we have economics on our side.

Compared to the cost of moving data, everything else is free.  As a result there are few who are not jumping at the opportunity to save serious money in their network.  When executives publicly state they see an opportunity to reduce CapEx dramatically and reduce OpEx by as much as 70%, people take notice.  Those expectations are buying signals which drive entrepreneurs to come up with new ideas, venture capitalists to fund those ideas, and buyers to look for start-ups with promising new capabilities.  As in any new market the risk for buyers is separating the wheat from the chaff.  Will Cisco dominate or go dormant?  Can an early leader like Nicera keep its leadership role having been acquired by a mass virtualization focused company like VMWare?  Can the start-ups survive long enough to bring the market the critical mass required to move us past the point of no return?  Will the big companies invest, partner, and acquire to expand their capabilities or to remove competitors from the market?

Although I cannot pick the winners and losers, I am confident all arrows are pointing in the same direction.  We need to keep the faith and foster the willingness to try, fail, and try again.  As with any technology the road to success will be paved by utter failures, however as long as the direction is toward a software driven world I expect our destination will be well worth the journey.