In these articles on enterprise class data centers, I have tried to bring a different perspective to successfully managing a data center. As I finish the final edit on a benchmark report exploring the area of IT Transformation, I am struck by what the data has revealed. The report uses data collected for approximately 200 enterprises. We specifically tuned this year's survey to include several questions about what factors contribute most to the success of IT departments. The Nemertes' benchmark report is nearly complete, but I want to share a key finding with you prior to publication.
Centralization
In short, the answer is “yes.” Centralization refers to services, like data center, and support, like data center operations. Of our respondents, 37.3 percent indicated that they were using centralized support to manage their IT infrastructure. Those respondents represented large and small companies as well as domestic and global. An additional 40.7 percent indicated that they used a combination of centralized and local support. In other words, 78 percent of those surveyed were using centralized management as a foundation for managing their infrastructure.
It doesn’t stop there. We also measured the type of IT culture in the enterprise. Cultures can range from conservative to aggressive. Conservative IT cultures do not leverage new technology or processes until later in the product life cycle. They are late adopters and will usually only do it when there are no options to maintain the installed base. Aggressive IT cultures are on the bleeding edge being the first to implement a new and promising technology to gain strategic advantage in the marketplace. The results of our data showed a strong correlation between centralization and the success of the IT department. Those IT organizations that leverage a centralized infrastructure are also the most aggressive in implementing new and evolving technologies.
From a data center perspective, the degree of centralization is reflected in consolidation of the services and applications into fewer instances in few physical locations. Centralization also includes colocation of the data center or outsourcing to a single location managed by a third party.
Who’s on First?
The surprising discovery that emerged from our data analysis shows another element that is an even stronger indicator of IT success. We measured organizations along several factors, including: clarity of roles, maturity of operating processes, consistency of architecture, automation of metrics, cost effectiveness and success in service delivery. Of all those factors the leading indicator of success was clarity of roles. Almost all of the respondents had centralized to some degree. What differentiated the ones who were very successful from the others was how well roles and responsibilities within the organization were defined. Centralization works best when all of the IT organization knows what their role is and what they are responsible for doing in any given situation. Of all of the factors we studied the role definition was also the one that lagged the furthest behind. Implementing new tools, defining technical architectures and detailing processes were all further along than the role definition in IT organizations that were not as successful as those who had a focus on clarity of roles.
We are technologists and are more comfortable dealing with technologies and tools. That is why addressing the complexity of role definition seems to lag behind even though it is the most powerful element we can use to improve success. Nor is clarity of roles as simple as writing a good job description. Each IT culture is a little different and supporting 50,000 users worldwide is different than supporting 500 users in a single site. Clear roles will eliminate confusion, improve service levels and mitigate operational heroics needed to resolve problems. All of those lead to a more satisfied set of end users.
To get more useful data center management strategies and insight from Nemertes Research download the Q2 Data Center Knowledge Guide to Enterprise Data Center Strategies – Volume 2.
