New Research Explains What Really Makes a Successful Hybrid Cloud

Another damn cloud image!

Seriously, how many new images of clouds can I post??

Anyone that knows me knows that I am a sucker for excellent research and data. Well, my friend and co-author on Visible Ops – Private Cloud, Kurt Milne (@kurtmilne on Twitter), just sucked me in with some excellent new research on ‘Private and Hybrid Cloud IaaS Project Success Factors‘. It was published by the IT Process Institute, so I knew immediately that it would be both interesting and important, and that it would provide immediately actionable insight – because that is what ITPI does. Indeed it is all I expected – 31 pages of pure awesomesauce, to be exact!

In it, ITPI looked at the success (or otherwise) of IaaS clouds from 143 different companies that have deployed a private or hybrid cloud. The report then evaluated a set of five key performance indicators – development agility, operational efficiency, service quality, business outcomes, and governance control – and identified the common features, functions, and capabilities shared by the most successful performers. All of this makes sense to me, as it reflects the key outcomes that most of my customers are looking for when they deploy or buy cloud infrastructure services.

There were some very interesting outcomes among the key success factors that the research identified. For example, it looked at the structure of the most successful IaaS environments, and found that the most successful performers chose cloud solutions that support a variety of virtualization technologies (i.e. not limited to single platform). Moreover, a majority of the top performers actually deploy their cloud across multiple environments too. Similarly, the top performers mixed their private and public cloud solutions to enable agile cloud delivery across multiple boundaries.

At first glance, much of this seems counter-intuitive, as many pundits advise adopting a homogeneous cloud platform to reduce complexity and cost. However, the research explains how this heterogeneous approach actually works in practice to allow a “best of breed” approach for multiple different workload types, technologies, and virtualization platforms, and to prevent another new single purpose or single technology silo. Smart thinking.

Another key outcome was that the most successful cloud deployments are highly visible and ensure executive support – as opposed to running a skunkworks project, or doing some small “under the radar” departmental deployments. When we look at new technologies, we often consider the benefits of developing grassroots skills and experience, and how to simply get the job done with minimal fuss. It is often intuitive to adopt these skunkworks programs, so that you avoid not just the hassle and bureaucracy, but also the large cost and risk of a highly visible project, while building this experience.

However, we also frequently see advice that talks instead about the importance of gaining high-level buy-in, but like a lot of common wisdom (which, remember, tends to be neither J) we have little data to support that assertion either. Now this new research finally answers the question quite definitively, with empirical data based on real world success. It turns out that executive – notably C-level – sponsorship is strongly correlated with cloud success, because it ensures that the effort is tied to business success and aligned with business objectives, not just another wild technology goose chase.

There is another heartening outcome that confirms what Kurt, Jeanne Morain, and I wrote about in Visible Ops – Private Cloud. The organizations that are more mature in their ability to standardize, optimize, and automate their infrastructure environments. Key technologies and solutions such application lifecycle management, automated virtual and cloud provisioning systems, configuration and patch management, and capacity management turned out to be key performance indicators too.

Again, as the research explains, this makes a lot of sense. A good understanding of standard processes and exception handling allows a more reliable and highly automated operating model, which is vital to the self-service and on-demand aspects of a successful cloud. It also confirms what we wrote in Visible Ops – Private Cloud about cloud success being heavily reliant on implementing high velocity for a set of automated “known good” processes.

There are many other incredibly interesting and useful data points in this new ITPI research, and a lot of very insightful analysis of that data too. I won’t spoil all the research for you anymore. The only other thing I will say is that you really should go and get the full study. And if you don’t want to buy the full research document just yet, or if you want to get a good sample of the sort of information you are missing out on, then you can take a sneak peek at the results in a free summary of the research, available on their Web site. The link to the whitepaper is here. And remember to follow @kurtmilne on Twitter!


This post was originally posted on’s Cloud Storm Chasers blog