Blog Post

Cloud is a corporate strategy, not a tactical solution

Stay on Top of Enterprise Technology Trends

Get updates impacting your industry from our GigaOm Research Community
Join the Community!

As an IT community we are still stuck in the past relative to the strategic nature of cloud. Many of us are looking at the adoption of cloud as just another technology, and are leaving the decisions on how to adopt, own, and manage the cloud up to engineers. But acquiring a cloud management platform is not an engineering decision — it’s a strategic one. Do engineers need to be involved? Yes, but your cloud adoption strategy has already failed if you don’t treat cloud as the operational construct that it is.

I wrote “Cloud management, what’s the big deal” a little over a year ago and the good news is many more of us now at least acknowledge the need for robust management tools. The problem is, we still think of them as “tools”. Cloud management isn’t just a pretty wrapper that you put on top of virtualization to make it easier to use, and it’s not a few scripts that automate builds or scaling functions. Cloud management is a platform that allows the cloud(s) owner to express their company’s directives and policies effectively and safely onto their myriad technology solutions and across international borders.

Why the cloud management platform you choose is so important

Like any software that solves a problem or creates an opportunity (often one and the same), a cloud management platform should be acquired only after defining a clear set of requirements. The requirements should be defined with the CIO and I’ve explained why after each requirement. A cloud management platform should:

Be capable of managing a variety of clouds – A strategic vision for where and how clouds will be adopted or dropped is important for a number of reasons: avoiding lock-in, the ability to retrieve data in a usable format, finding the appropriate cloud platform for the expected workload and location. Depending on the business you’re in you may use partnerships and or competitive concerns as a decision factor in your multi-cloud strategy.

Handle data security and location – Do your systems administrators have access to corporate strategy around locations and data privacy requirements? What about HIPAA or other regulatory concerns? If they don’t even recognize this as an area of concern, why would they look for it in a management tool?

Take care of policy management across clouds – Your architects and engineers might be terrific, but are you sure they are the best ones to determine the value of having a common and simplified set of tools for managing policy and governance across your images and across different clouds? Policy considerations can take into account everything from privacy to security, to performance and lifecycle depending on the platform you choose.

Include well-developed role-based security – While your engineers and infrastructure leaders are more than capable of handling security decisions for team access to a cloud management platform, are they the right group to determine how customers (developer or end-user) and partners might access your cloud?

Incorporate a virtual machine security suite – This is an area where the CIO likely doesn’t need much involvement, but there should be a senior security role involved in the project.

Consider the full life cycle from creation to deletion – Unfortunately, most of us in the trenches don’t think about whether the images we create today should be reviewed six months from now? Ensuring you have a solid life cycle approach will help you develop a more efficient use pattern and reduce the risk of inappropriate resource use.

Integrate with operations platforms (monitoring, billing, etc.) – The effort to define these requirements will mostly fall on the technical team, but feedback from management about expectations of monitoring and billing etc., is still critical.

Offer APIs for common tools and scripting languages – Mostly a technical/architectural decision, with the exception of integration that might enable out of the box opportunity, there might be value in having a larger team, including leadership involved.

The above isn’t a complete list of considerations in the evaluation of an appropriate cloud management platform, just serious food for thought. However, of what’s missing above, the most critical element of all includes thinking about how a cloud management platform should complement and re-orient your IT organization.

Where the rubber meets the road!

The CIO needs to consider a ground up redo of the organization, how it delivers IT services and how it integrates with the business at the function and end-user customer level. Admit it; you weren’t thinking that these organizational changes were a factor in your requirements and prioritization process for acquiring a cloud management platform.

Of course, the aforementioned needs aren’t necessary if all you want is a shiny set of tools or some home-grown scripts to handle your cloud. Tools and scripts which won’t scale, aren’t standardized, won’t work across clouds and will likely be developed differently by each IT group in your enterprise.

The delivery of IT is different. The old ways are gone. The addition of cloud to your organization isn’t an opportunity to do the same old things faster, it’s an opportunity to deliver functionally improved IT services in real-time to your business.

How will you deliver in real-time if you still work through a traditional helpdesk process? Or maybe because you haven’t figured out how to integrate billing, you still have business or IT groups who want to “pay” for their servers. What about the purchasing process or the approval process for a new application?

This isn’t just more tech bells and whistles

In modern IT you should be able to test, fail, test, fail, test, and implement in less time and for less money than one effort in the past. In light of the improved application adoption options, a change in how you review and approve ideas is also important. In other words, why send a project to the executive team for review, when you could run a proof of concept in a matter of days or hours and actually demonstrate the value almost immediately at little or no cost.

OK, maybe there are more bells and whistles, but treating cloud like a technology solution purchase is the wrong approach. Take a holistic approach to how IT can and should participate in the business of doing whatever your company does, then build the operational model to support that. Seek alignment in your organization and in your technology choices so you’re prepared for a Cloud Operating model and Fluid IT. Welcome to the modern IT world.

Mark Thiele is executive VP of Data Center Tech at Switch, the operator of the SuperNAP data center in Las Vegas. Thiele blogs at SwitchScribe and at Data Center Pulse, where is also president and founder. .He can be found on Twitter at @mthiele10.

93 Responses to “Cloud is a corporate strategy, not a tactical solution”

  1. Eugene

    I am a great resource to transfer your IT platform to the cloud ,making your life easier and more efficient. Send me a message so I can tell you more .

  2. Hi Mark, interessting piece… I was triggered by the headline: Cloud is a corporate strategy, not a tactical solution. Beacause thats what we say about CRM. I’m organising a meeting for a group of 100-150 local medior company owners at sept. 13th about The Cloud. In advance I wright a piece in the local magazine about it and I’d like to use some of your information in there, offcourse mention you as author. I also shared it already on LinkedIn and Twitter. So thanks you gave me a good picture of your ideas about Cloud. Best regards, Petra Hagemeijer, GM/part owner at Zinnovation, Hoorn, The Netherlands.
    PS, your name sounds Dutch…is that correct?

  3. When I suggested in several Linked In discussion groups that Cloud Computing was a strategic information technology infrastructure logistical technology which was being driven by the mobile smart device revolution I was ostracized. What’s really funny is I use cloud computing applications on my laptop more than I do on my smart phone. While my wife barely touches her laptop anymore since she got her iPad.

      • Hi Mark,

        Thanks … But it is no longer a fight! CIO’s of major corporations are beginning to recognize that Cloud Computing has become the primary IT Infrastructure Technology evolution for 2012 and throughout this decade. Big Data; and ETL, Data Integration and Aggregation technologies will feed the Business Analytic and Intelligence technologies as their evolution will evolve with Cloud Computing technologies.

  4. Sean Patrick Gallaty

    It is a delusion to think that any strategic decision is solely one of management. Cloud is a tool, just as thin client, client/server, 3 layer, SAS are all tools. Every year has it’s hallmark groundbreaking new thing. There never has been, and never will be a reason to eliminate the technical decision makers from the process no matter how abstracted the solution.

    • Sean, I’m not sure you read past the title of my blog. Nowhere in my blog does it say “technical staff shouldn’t be involved”. However, I must disagree with you on Cloud being a tool. It’s only a tool if you’re considering it as piece of a larger objective. Cloud management must have strategic vision applied for you to both benefit from the new opportunities and cloud and protect yourself as well.

  5. Goli Afshari

    Great article. I have worked for HyperOffice for many years and I have helped hundreds of clients move their data to the cloud. It doesn’t lock them in, it is secure and roles are defined. It has the most powerful role-based security I’ve seen. It also has a database which is really powerful.
    I agree with you in that users have to change their mind set… you have to let go of the traditional help desk model and work in real time. In HyperOffice, I help companies create their own landing pages and design the UI of what their employees and clients see. I think, flexibility in the cloud tool is key in helping traditional users make the move to the cloud.

  6. Gerry Grealish

    Great article Mark. Reblogged for visitors at PerspecSys and commented:

    While, as you say, it is not a complete checklist, it is a heck of a good starting point.

    Cloud is not simply a set of new “delivery pipes” with some TCO benefits, it is a transformational new approach that empowers business units within the enterprise. But CIO’s and CISO’s are responsible for ensuring compliant, efficient and effective use of it, without hampering the innovation it promises to unleash inside the organization.
    Multi-Cloud, full life cycle capabilities, security support for varied access/use, plus the ability to adapt to the shifting sands of international privacy law/sector compliance requirements…all critically important stuff. I’ve been seeing the last area come up more and more over the past 6 months – IT and Security organizations in Germany, the UK, Switzerland, Australia, Canada and others taking steps to design capabilities into their cloud management platforms to keep sensitive information out of multiple clouds, kept resident within their own firewalls. Just another example of how CIO’s and their teams can meet the needs of the business lines (who are hungry to adopt cloud) without trading off on security/compliance.

  7. Thanks for the fresh perspective on cloud management! I am in the process of writing a series on cloud implementation for businesses and I will definitely be referring back to this!

  8. Vishal

    God article which focuses on strategy of cloud. I think the point which really matters is cloud is not about doing same old things faster, as commonly perceived by people, and this article makes a very good point on that!

  9. Chris Leigh-Currill

    As ever a great article. I whole heartedly agree with your message and it is very close to what we are driving at over at Ospero. We are especially aware of the considerations to data protection and location issues for businesses operating inside Europe.
    However one observation I would like to make is that in many businesses they have the additional challenge of the ingrained siloed development culture, very prevalent in banks, where there are lots of big concurrent IT projects reporting to different people with budget. Each one of these projects will make the decisions best for project delivery and this will lead to many different technologies being used. For example I have worked in places where one project has chosen Business Objects for their BI tool and another MS Reporting services, both decisions were correct for the project but which one was right for the business? This is especially hard as both technologies were being used and so either was a valid choice. I fear will be no different to cloud strategy as many businesses will initially take a toe in the water approach with a couple of projects. With little or no thought to a long term strategy. If a business cannot get its BI strategy right, a concept that has been around for over 20 years now then what hope the cloud.
    As you say business should look at the cloud not as just another technology to add to its no doubt long and over bloated list but as a chance to look inward, evaluate what it is IT is doing for the business and embrace the change. This will require a CIO with tremendous will, understanding of cloud, agile IT, political agility and probably compromising photo’s of the CEO and CFO ;-)
    The unfortunate consequence is that the businesses that stand to gain most from implementing a coherent cloud strategy will find it the hardest to implement and I believe that most will initially fail.
    Chris Leigh-Currill

    • Chris, It’s a great point about the Silo’s. That should be a strong leaders first objective in IT to break those down. Unfortunately, many CIOs fail to make the change they would like because they are assumed to be “short-timers” by the rest of the team. The team leverages the short timer status to delay objectives in the hopes the CIO will be gone before any work actually occurs (Status quo is maintained).

  10. Mark,

    Insightful article. One area you touch on is the importance of the whole life-cycle. I find many people don’t appreciate how much of an impact the amount of time it takes on the life-cycle front-end (from approval to in production) and on the back end (from the deletion decision to getting it off the books) has on the cost of a project. Whether it is hardware or virtual, with today’s shortened useful life, two weeks or more here and there can have a double-digit percentage impact on project cost.

    Effective management systems and processes should be able to reduce this.

  11. A well written article, thanks for the contribution.

    In truth all major technology shifts merit being reviewed from a perspective of more than just the pure technology itself. Any technology of value gains that value from the fact it enhances or impacts the way we work and what we are able to achieve. When those basic elements are altered it is necessary to consider the impact across the organization.

    Cloud computing is more than just simple – more/better/faster – it alters the style, reach and depth of collaborative work between people.

  12. Dennis Kilian

    I like the article. I was sort of curious about your last paragragh. Are you saying that IT should be able to develop POCs faster as a result of adopting a cloud strategy? I am also curious about the idea of developing a POC prior to taking it to the executive team. Test what? Implement what? I understand the value of a POC, but developing a POC based on what, if not input from the executives who interface with different stakeholders?

    • What I’m trying to say regarding POCs is simply that with simplified resource creation/deletion IT is in a better spot to search for alternatives to current tech or find new solutions and test them in a low risk environment. This testing would be regardless of where the idea/need was generated from. What I’m not saying is that all ideas should come from the business or vice versa.

  13. Just Visiting

    Interesting perspective… but it falls short of a strategic position. Everything in your article is correct, but its simply the symptom of a more fundamental issues.

    CIOs are right, cloud is an extension of their standard portfolio of hammers, screws and nails. How can anyone expect a CIO to take a strategic view when cloud products are packaged and sold as rivets? The “real” value behind cloud computing is challenging to achieve.

    There are several reasons,, Analysts and bloggers focus on hype more than the less glamorous meat and potatoes infrastructures that delivers core value.

    Companies (outside of the telcos), accomplished infrastructure providers, fund meaningless open technologies with chest beating, and a feeling of self-wonderfulness with the hope of good will and accelerating adoption. Too bad open technologies just create more screws and nails and not value. Any initiatives building value delivery, goes unfunded.

    SDOs, mutual benefit standards organizations, exist for the benefit of their members, mostly manufactures and global integrator. They only invest in works that better their positions in the market, not the customer’s.

    CIOs and staff are not usually part of those clubs and don’t contribute to their futures. There an saying “You can’t complain about the election if you don’t vote”

    Cloud will always be considered a tactical solution because it is a tactical product !! Along with all the other silo’d point solutions.

    The strategic value needed today is not IT management, cloud computing, virtualization or any other of a million point solutions.

    Value, core value is in a strategic delivery infrastructure. It doesn’t exist today.. Not because we don’t know how to build it.. Its not in our hands because it goes unfunded. Why ?

    Most of the technologies are challenging to orchestrate and choreograph. Mid-range providers cannot afford to build it themselves and must move to a funded open source model. Funding requires partnerships.

    If the partnerships become mutually benefit foundations, its no better than the SDOs. Open, customer focused initiatives are unfunded because it reduces consulting revenues, further commoditizes virtualization technologies, providers that can benefit are busy chasing hype and beneficiaries take a passive roles hoping the ‘market” will magically help them and don’t contribute to initiatives that reduce their costs.

    One hope are the telcos, if they can move quick enough. It won’t be open source, and with the proper guidance, it will get the job done.

    Yeah, its a hard nut to swallow.

  14. Devi Mazumdar

    It is disturbing to see how many businesses move their entire infrastructure to the cloud irrespective of what their business needs dictate. As IT solutions provider, our job is to “prescribe the right medicine to the patient”, while guiding them to a healthier and happier future through strategic technology deployment, informed by early adopters.

  15. David

    Maybe its just me, but if you own your cloud, then we are really talking about your corporate network. It seems like the real challenge of ‘the cloud’ is that fortune 3000 is buying hosted services outside their firewall. Am I missing something or was ‘cloud’ such a cool word that people had to pull it back inside the firewall so they could have fun too.

    • David, The issue isn’t whether a cloud is internal or external, but rather that cloud technology while important in what it enables, is just a technology. The opportunity is in how the organization as a whole leverages cloud. Leveraging cloud is much more than some added efficiency or scale, it’s people, speed to market, M&A, new products & services, organizational alignment and much more. Lastly, the likelihood that any business will be using just one cloud (public or private) is very small, unless they are an SMB, so a strong management strategy is necessary to guide in a common fashion how all the clouds are utilized and secured, etc., etc.

      • David

        I’m just lamenting the loss of the term. ‘Cloud’ could have meant ‘hosted by a third party outside our firewall’ and the risk/reward trade-offs that come with that. If ‘cloud’ just means ‘not on my local machine’, you are correct in asserting that it is just a technology. In fact it is essentially the same client server, portal and networking technology that we have had for decades before the term ‘cloud’ was coined. The challenge is that we will need a new term to describe what companies are taking on when they allow third parties to host critical aspects of their technology services and proprietary data outside of their own firewalls.

  16. Michael Hoffman, Inc

    I appreciate the candor of your blog. Many of the cloud salesmen I see blogging these days are pushing the spiked kool-aid on executives rather than addressing an actual problem. As you mentioned, the considerations you noted are not all-encompassing, but I was surprised not to see the implications of taxes for moving to the cloud. What are your thoughts on this topic? From what I have read, there are state to state implications that may impact the move towards cloud. SLA is the other obvious consideration that I feel needs to be looked at it with a pessimistic eye. There have been many failures to date, the American Eagle commerce site being the most visible for me. Any failures of this magnitude can make or break a CIO’s career or put you at significant risk of legal action. I wholeheartedly agree that a move towards cloud as an IT strategy rather than a technology requires direction from the top down. It is just going to take time to adopt. I look at cloud much like I looked at SOA. Eight years ago, I had one salesman after another coming to my executives telling them how they had to redo all their systems to be service based. It took time, patterns evolved, products matured and now SOA has become the norm rather than the exception in product development.

    • Hi Michael, thanks for the comments, much appreciated. As you mentioned, I didn’t attempt to cover every concern. Legal & taxation risks are two serious areas of opportunity and concern. Generically, if you’re using public cloud I think taxation issues are minimal. However, legal issues around data location, recover-ability, HIPAA, & other regulatory concerns is real. Happy to chat more if interested.

      • Chandrashekhar Tanwani

        Thanks for thought provoking article. If we look back to the SOA and EA evolution, though these were Technology related concepts, these impacted the business and operation model of whole Enterprise, and so required the attention of Business Folks and CxOs as well. But with ‘Cloud’ that is not the case. It does affect the operations of IT department alone and not the whole Organization. The change can be transparent to Business or End customers without major ‘change management’ issues. Though ‘Cloud’ imlementation raises some legal and security issues but that means the CIOs need to learn these aspects also and some new policy measures need to be defined but it still can remain ‘IT only’ initiative with economic benfits passed to the business.

  17. David A. Smith

    Very good article!

    It highlights that companies need to incorporate a suite of cloud solutions to remain competitive. He states, “The CIO needs to consider a ground up redo of the organization, how it delivers IT services and how it integrates with the business at the function and end-user customer level.”

    He also emphasizes the need for continuous service improvement with his comments, “In modern IT you should be able to test, fail, test, fail, test, and implement in less time and for less money than one effort in the past. In light of the improved application adoption options, a change in how you review and approve ideas is also important. In other words, why send a project to the executive team for review, when you could run a proof of concept in a matter of days or hours and actually demonstrate the value almost immediately at little or no cost.”

    • Sophocles

      Dave..Good comments…. As a 45 year veteran of IT in the trenches,experiencing several “evolutions” of technology, I would be skeptical of any “proof of concept in a matter of days or hours that actually demonstrates the value almost immediately at little or no cost.” My experience with “Proof of concept” has been pretty good when used to test validity for the technical aspects, but falls short when attempting to extrapolate the validity relationship to the anticipated new business system! To assure a minimum of problems, delays, and restarts, Be sure that the system analysts, system analysts, and programmer team leaders employees are current employees who have been with your firm a minimum of two years (preferably three to five years). Put extra effort and time into developing the system specifications (emphasizing in detail the new system processes with sign offs with all affected people).
      Do not authorize ANY programming until the new business system has been reviewed and approved by all affected parties. Extra time applied during the design phase pays off handsomely during the programming and testing phase! You MIGHT bring up a system without doing these things, but I can assure you, the system will be not be on schedule, will be plagued by “bugs” and false starts, and will be over budget even with a topnotch technical staff. System excellence begins and ends with the system design!

  18. Good article. However, I felt most of the points used in the article were technical arguments/points used to support a management vision: that cloud is not a tactical solution, it’s part of the corporate strategy. In that sense, probably points like cost, elasticity of service and other management pros/const would have been more appropriate.
    In the current format, it does feel like the article is addressing to a converted technical audience.

    • I Am OnDemand

      Mike – Good point, though I find it valuable as well to explain what actions and considerations should be taken in order to support support strategic plans.

      • Steve Hammond

        I have to agree with Mike. The issue of governance even in the cloud shouldn’t be ignored by CIO. The difference of cloud infrastructure to traditional infrastructure for upper management is being able to manage real-time operating costs more closely. But let the butcher trim the fat if you want lean meat… choosing which platform, API, security, and overall implementation will most likely be handled by the people who manage such now (as they can estimate the friction involved more closely)

  19. Mayuresh Jehurkar

    Great Article! Thanks for the insights!
    With innovative platforms, like ComputeNext(, coming up it definitely makes life easier for a comparative analysis for collaborative decision making by the management and technical teams for purchasing “right resources @ right price”.
    Auditing, Billing, Resource management, Tracking probably covers all.
    However, interesting thing to watch would be how the industry behaves to this new wave .. coz this wave is taking it up in the “Cloud”

  20. Don Turnblade, MBA, MS, CISSP

    Also, the legal controls such as forensic images of a cloud incident by unrelated parties that just happens to capture your data in the image. Do those lawyers even know your data needs special compliance treatments?

  21. Alex Jauch

    Excellent article, thanks!

    Totally agree that a technical approach to Private Cloud is too limited to drive successful cloud deployments.

    This is one reason why I wrote “Why we Fail” which is focused on the business model changes you need to make when deploying private cloud:


  22. I Am OnDemand

    Mark – Thanks ! great piece (as always). AS you mentioned these requirements are not the only factors for getting yourself the right Cloud Management solution – one important point that I would have add is the > SLA auditing and measurement.

    Cloud adoption must come with clear understanding on how the new environment will be managed. As you mentioned, CIOs that see cloud as an evolution use their traditional IT tools and expertise to control their new bulk of endless virtual resources. Eventually they find themselves lost when the “surprising” high cost alerts (the new monitoring ?! :) ).