Welcome 2014! I hope it is a glorious year in which we all realize our healthy doses of the tremendous change that has been bearing down on us for the last couple of years. What change? You know the line up: cloud, big data, real-time computing, mobile, etc. For 2014,with the change, we will also continue to see a fair dose of hype. The hype will go on and possibly increase in volume as vendor providers aggressively click their hockey sticks at the puck of progress. For this blog, I’d like to get into one topic I feel is getting over-played at the moment; public clouds.
No haters please! I am not saying public clouds in total are hype, but there is an expansiveness that reads to me as overly eager. Public clouds are a handful of years in the making and there are some great successes to point out “Salesforce”, “iCloud”, “Ariba” are just a few examples of successful business models that have proven the validity of public clouds. These examples have additionally demonstrated adequate stability for businesses in order for companies to consider leveraging this innovation. These are rudimentary steps toward an inevitable shift towards what I call the utilization of IT. Public clouds will continue to increase in importance. The ultimate spoiler was Nicholas Carr when he called out this trend years ago in his book “The Big Switch”. I also agree this is not a question of “if” but “when” and in what manner. With all this said, I think we’re going too fast. I am seeing an over-rotation to what looks like a “2-minute drill” to drive public clouds into all aspects of IT for mid market and multi-national organizations, ASAP! However, I believe we’re still building the necessary groundwork to make this feasible and I worry about the risks.
One doesn’t have to watch the news for more than a couple of weeks to see a significant cloud outage, or massive breaches in corporate security. As a global industry that is building new cloud disciplines, we are less than a decade into a massive inter-connected world where everyone leaves their digital signature throughout their endless online activities. Corporations open themselves up to constant attack from malicious opportunists in order to try to keep up with the growing opportunity within this global online economy. This alone is a MASSIVE step forward and we haven’t gotten it right yet. Like a fine wine, our processes and technology have to evolve/mature. We’ve become a society where the gas pedal must stay flat to the floor (and I’m the worst…), but we have trillions of dollars at stake in this evolution. If we have learned from our history (e.g. DotCom, Year 2000 bug) we will remember as IT escalates and speeds beyond our ability to consume the change, we find ourselves sub-optimized and less wealthy from our over-steps. Ultimately in these scenarios we find ourselves led to a correction of some sort and back to the whiteboard for remediation work. It’s a significant hit to productivity and unnecessary.
Evolution, it takes time…
If we all agree it would be great to avoid another unproductive, over-zealous dive into the future, let me offer a governing function for your consideration. What I see trending in businesses is a smaller, more prudent step that creates less risk, but still delivers profound change that optimizes operations and builds a foundation for better decision-making.
I call it the Option E business model where the “E” stands for everything. If given a clean choice, I believe corporations will not choose: “Public Clouds”, “Private Clouds”, “Appliances”, “Traditional Data Centers”, but option E… all of the above. Companies will choose to leverage all types of market offerings in order to address their specific needs and strike a balance between value and risk to match their profile. This approach can be aligned with the concept of the “Software Defined Enterprise” (SDE). SDE is the transformational change where functionality is abstracted away from its hardware limits through virtualization. We’ve all tested the waters of virtualization and reaped the benefits. However once you are 100% virtualized and you move into this realm of SDE, it’s like crossing over into OZ. Now things that were manual and slow, can become automated and agile. No longer do you have to wait on a FedEx truck to stand up a replacement server. You can provision automated/on the fly, and disaster recovery becomes a button click. This is all possible because now it is virtual: a software-based world.
What’s important here is SDE is integral in the delivery of all clouds, so you will be leveraging SDE regardless of your plan. The question is will you get the most from it? Are you reworking your processes to exploit it? Have you squeezed out the buffers and slack time that are prevalent in traditional IT processes? That “protective layer of fat” was there to reduce risk, but now it is not needed. Ultimately these efficiencies equate to real dollars and speed to the business.
Ok, so why not just jump to public cloud and wash your hands of it all? Aren’t public clouds the future? Maybe, but there are some good reasons to gauge your plans. I already mentioned there are more risks associated with public clouds than your traditional IT operations. I want to add a few other reasons to consider. A big reason you should pace yourself is cost. Today you are evaluating the “cost” and “speed” of public clouds against your traditional operations. This is an unfair fight. You have potentially $10-$50-$100 million or more of optimization opportunity within your business that can be easily attained. One of the best examples of this is EMC’s transformation to SAP and ITaaS. The project was called Propel (http://bit.ly/LBARIA), please give it a read.
There are many examples of companies insourcing workloads that previously sat on public clouds, and as a result saved millions in operational costs. Publicly offered clouds are first and foremost convenient, although they rarely stack up against a hybrid SDE model on protection, cost and flexibility.
Finally, some of the public cloud offerings today are a bit like Hotel California: “You can check out anytime you like, but you can never leave….” Today, It can be difficult to divest from a public cloud decision. Companies who have spent little to no time building disciplines and processes around SDE ( service brokering, automation, classification, portability, migration, security, etc) are taking undue risks with their customers’ data and with their corporate viability itself. We must remember that, although as a global commercial/industrial complex, we are working to reduce the cost and complexity of Information Technology, we have never been more DEPENDENT UPON IT!
In summary, its time to evolve, plan your steps wisely. Look to convert IT from an operations function, to a broker of services leveraging all that is available to you to meet and exceed your constituents’ expectations. This means developing news skills and new processes. Also it is important, to keep your C’s & V’s from signing quick checks that rob the company of opportunity and take away the CIO’s ability to effect positive change. Build a story about your approach to evolving the business. Taking control of that corporate perspective of clouds is your prime objective. It’s not really about clouds, it’s about providing a service. Don’t let your company give away your competitive advantage in an effort to evolve. Good luck, and ping me if you want to brainstorm about your company’s plans.
Andy Sitison 
Andy, great points here! Fully agree with your perspective.
Let me further add my person experience and perspective on this.
Note that most customers have a policy (even if informal) to prevent “vendor lock-in” where they split their needs across vendors. One key lesson learned from the big outsourcing contracts from the last decade was: define the exit clauses before getting in!
Also, industrial espionage episodes as well as episodes of cloud service providers going out of business or at least temporarily out of service have happened, and further builds the case for choice.
Indeed, many companies have been moving back in (in-sourcing) systems once on public providers. I’ve seen multiple reasons, but one relevant one has been the lack of flexibility those cloud “industrialized” solutions offered.
Not to mention the costs they come to represent when considering highly active and critical systems! Having worked in my past life for an outsourcer, I’ve realized one simple thing: the outsourcer aims for profit. So, we are usually talking about at least 30% gross margin. When considering internal IT, it is a cost center, so not margin on top.
Well, you need to get lots of operational efficiencies to provide the same service at at least 30% less cost! and let me tell you from experience, many haven’t made it.
The analogy regarding what is called today the public cloud, was for many years the energy industry: why produce your own energy when you can get it from a power plant?
But the fact is that this market is again changing towards smart grids and micro-production bringing the energy production to people’s homes.
Also, most critical entities often consider own production capacity. Look at the most modern datacenters, many include own generation capacity – not only for the case of network power failures, but also for regular production through renewable sources for further autonomy and less overall costs.
So, fully agree with you, that all hypes must be considered with caution, and that the key is to provide choice.
I believe that as we sail through this sea of change, before we reach that future you refer, there is an outstanding opportunity to help internal IT departments modernize and achieve significant levels of “industrialization” of their operations, which can only be achieved through software led automation.
It’s the elastic theory: options stretch and shrink in cycles, as the market tends to eliminate monopolies, as soon as one thing gets dominant, lots of providers will emerge to build better, and more efficient alternatives.
Public cloud is there, will further mature and see an increased adoption. No question about it. But in the coming years it will not be the only solution for all problems.
Maybe an hybrid model between a public cloud and a “hosted managed private cloud” would present the best of both worlds!
Thanks for your detailed response. Several great points
Andy,
On your related topic of public clouds, here is an interesting article and strong words from Seattle marketing technology company, Moz, and their struggles with public cloud. http://www.xconomy.com/seattle/2014/01/30/moz-dumps-amazon-web-services-citing-expense-and-lacking-service/
Good to see a publicly documented version of this frequent story. Thanks for being my reference section 🙂