The End of Private Cloud – 5 Stages of Loss and Grief

It’s not today, or tomorrow, but sometime in the not too distant future the bulk of the on-premise private cloud market is going to shrivel into a little raisin and die. A very small number of very large companies will operate private clouds that will be, by an large, poor substitutes for the services available in public clouds. However, they will be good enough for these companies for some percentage of their workloads.

I have seen dozens of private cloud efforts by many large customers. Most are pretty weak shells of a cloud, not coming close to the economics or capabilities of even 2nd or 3rd tier public clouds. Comparing them to AWS, Azure or Google is like comparing my art work to a Picasso or Rembrandt. The only similarity is that I can still call mine art even if it’s atrocious. I can still call your cloud a cloud too – even if it’s expensive, inelastic, and lacking anything but the most basic of features. Some will be reasonable, but in the long run it’s a game you cannot win.

No matter how good you think you are, you’ll never have the resources, skills or need to be as good as Amazon. AWS deploys enough computing capacity every day to run when it was a $7B online retailer. How many servers will you rack and stack today? How many petabytes of storage will you deploy this weekend? How many features did you update this year (Hint – in just the first half of November, Amazon announced 27 enhancements, features or entirely new services!!!!).

In her seminal work, “On Death and Dying,” Elisabeth Kübler-Ross articulated the 5 Stages of Loss and Grief. I think it’s time to look at this for private clouds.

1. Denial and Isolation

Most large company IT organizations are in severe denial about that is going on in the public cloud market today. They think that if only they get their vCloud or OpenStack cloud up and running they can be just like Amazon. Or perhaps they still cling to the total fantasy that their internal data centers are somehow more secure than Amazon’s or Microsoft’s – companies that spend more money on InfoSec per day than most enterprises will invest over the next 5 years. The denial comes from a fear of change, fear of loss of position and career, or just ignorance. By the way – denial is a guaranty that the risk is real. Those who see the future have already shifted their careers to ride the wave instead of being destroyed by it.

2. Anger

One you start to understand what is happening, that your career plans and worldview are being overtaken by the cloud, it is natural to become angry and bitter. You’ll go out of your way to point out potential security or performance issues with public clouds, maybe blogging about the “what ifs” of outages and disruptions, attack vectors and dirty power grids. You can’t control this because your CEO is not going to give you $100M it will take to really build a private cloud. Or perhaps you’re a private cloud vendor looking for that exit that may never come. “Oh why did I waste my time on this market” you might cry when all of the exits have passed you by and you’re looking at an ever dwindling market with lots of dying startups trying to consume whatever oxygen is left. Perhaps you can jump on (off) the Cloud Liberation Freedom Front (aka “CLiFF”)…

3. Bargaining

The normal reaction to feelings of helplessness and vulnerability is often a need to regain control–

  • If only I had moved to public cloud sooner…
  • If only I had gotten better advice from IBM, HP, VMware, Oracle or Accenture…
  • If only we had tried to be more cloudy in our data center…

Secretly, we may make a deal with our higher power in an attempt to postpone the inevitable. This is a weaker line of defense to protect us from the painful reality. Do we pray for AWS to fail? Do we pray for a Google data center meltdown?

4. Depression

It’s over. Sadness and regret set in and we realize that there is nothing to be done. Our best laid plans are in ruins. The future looks bleak, where servers are getting older by the minute, turning off one by one in silent desolation. The staffing model for 2020 shows a drawdown to a skeleton crew just keeping alive the old legacy stuff that you can’t kill or migrate. It’s dull, sad, drudgery.

5. Acceptance

Not everyone will get here. Many have already, coming to the early conclusion that the future is and will be in the public clouds. Those that do get here before everybody else will have more opportunity, more reward, more fulfillment. The late arrivals may have to find other careers – like today’s laid-off mainframe programmer looking for a job at Facebook, it ain’t gonna happen dude. Many a former techie has found fulfillment and happiness in other fields – I even know one who went back to medical school and is a practicing oncologist. Pretty cool, eh? Even Julia Child didn’t start cooking until she was 50 – so your second career is nothing to fear!

In any event, once you understand that the public cloud is the future – and when you are over the denial, anger, bargaining and depression – you can start to make plans.

CIOs should start getting ahead of the curve, thinking very hard about whether or not that new data center plan is worth the investment. Why spend $30 million, or $300 million, on a fancy new data center that may lay empty in a decade?  Instead of investing $5m in a new private cloud, how about investing $5m in the InfoSec upgrades required to safely use a public cloud?

It’s only a matter of time. Resistance is indeed futile. The public cloud is the future.

(c) 2013 John Treadway / CloudBzz / TechBzz Media, LLC. All rights reserved. This post originally appeared at You can follow me on Twitter @CloudBzz.


10 thoughts on “The End of Private Cloud – 5 Stages of Loss and Grief

  1. Hah. This is exactly how I feel coming back from re:Invent.

    I think private cloud technologies currently are great and exciting – OpenStack has very much matured. But during the re:Invent keynotes learning the sheer amount of innovation and features that AWS continues to deliver and other public cloud technologies plan to deliver I felt less excited about the relevance of setting up private clouds.

    Some big companies will need private installations – but for most companies, using public cloud providers and understanding how to build on top of them and controlling costs is a better approach long term IMHO.

  2. First, I don’t really agree with the numbers in this report. But assuming that it’s true, that’s today’s view. As Wayne Gretzky said “I skate to where the puck is going to be, not where it has been.”

    This post is about the long view. People think they will use private clouds, and many will, but in the long run few private clouds have a chance of competing with the large public clouds on features, pricing or usability.

  3. Moving away from on-premise does not automatically mean customers will move away from private cloud. Like many, this article confuses physicality with consumption. A private cloud is one in which there is only one tenant, but that tenant can exist anywhere in anyone’s data center. A public cloud entails multiple non-related tenants. I can see many businesses getting out of owning their own data center, but I don’t agree that the majority will end up in a public cloud as I have defined here for many reasons among them consistency, availability and, frankly, we see no movement by telecom providers to reduce costs for bandwidth required to operate at enterprise scale.

  4. JP – fair point on hosted private cloud vs. internal private cloud. That said, the vast majority of hosted private cloud solutions today are barely beyond basic compute, storage and network. Many still have manual processes in the back that limit their economics and scalability. The portals are weak, APIs are nonexistent or minimalist. Service catalogs are thin, etc. There are exceptions, but by and large hosted private clouds are only slightly better than the current state of internal clouds. It’s largely a transition market. Customers may move from internal to hosted private but an increasing amount of workloads will move to public or virtual private.

  5. I completely agree John.
    My team spent the last year moving all of UCAS’s (Universities & College Admission Service) mission critical IT systems out of an aged (end of life) on premise DC and on to public cloud )AWS and Azure).

    As part of that move we evaluated a number of the leading Private Cloud offerings at both a technical and commercial level and on almost every measured criteria they fell way short of the public offering. Our move was a huge success and delivered a phenomenal outcome for the business, it’s members and the applicants.

    I do however still have a chuckle at the enterprising salesman that offered me ” On Premise Private Cloud”.

  6. I found this interesting, it went well with what I am learning in my Psych 101 class right now. We have learned the five stages of a loss, in which this article has defined Elisabeth Kubler-Ross stages of dying very well. This stage of emotions come from a death of someone close to you or even yourself, everyone goes through it. Elisabeth interviewed hundreds of terminally ill patients and found they were all going through each step that comes with a death. Although death is a part of life each individual reacts to dying differently.

Comments are closed.

Create a free website or blog at

Up ↑

%d bloggers like this: