By Sramana Mitra and guest author Saurabh Mallik
This interview is part of our series on Thought leaders in Cloud Computing. I am talking to Mark White of Deloitte. Mark is the chief technology officer of Deloitte Consulting LLP’s technology practice and the lead IT principal on the Department of Homeland Security account. In his six years at Deloitte, Mark has served clients in the federal government, financial services, real estate, high tech, energy, transportation, and other industries. Guest author Narayanan Raman interviewed him last year as part of the CIO Priorities series.
SM: As a thought leader for your company, where is your sense of cloud computing adoption today?
MW: It actually has progressed well. We spoke nine months ago; part of the discussion was about what we see CIOs doing based on the economic downturn and the potential back up. I think at that time one of the topics we discussed was how the economic downturn accelerated the evolution of cloud by at least two or three years, because in many cases it forced experimentation and adoption earlier than we would have otherwise seen.
Because of that, with the type of work we are doing with clients now, we have definitely moved farther along the path of adoption toward maturity. I like the Gartner model, which has placed cloud at the peak of inflated expectation. Let’s take a year ago, where the vast majority of work we did with clients is what we call phase 1 or type 1, which is to understand the framework of the cloud, to have actionable conversations about the cloud, and to understand the potential and the risk of the cloud. So, it is very much about trying to understand the framework. Now, type 2 work is the actual planning, strategy and business case, design, and selection, which includes all of the work we do leading up to implementation or adoption. In type 2, I would include experimentation or putting your toe into the water. In type 3, we have more implementation and rollout. So, a year ago, it was 90% type 1 work, 8% type 2 work, and 2% type 3 work. Today, a year later, it’s probably still 20%–30% type 1, talking about and framing the potential and the options and understanding the risks, the obstacles, and enablers. There is 40%–50% type 2 work, which is actual projects involving planning, strategy, design, selection, and preparation for implementation, including a significant proportion in gate activities. The remaining 30% is in larger scale implementation.
SM: In larger scale implementation, what workloads are going into full-scale deployment right now?
MW: Kudos, first of all, for using workloads as the noun. That is the right thing; we need to ask ourselves what workload to adopt. Now, to answer your question, I would say that the vast majorities still lie at the edge of the enterprise. By edge, I don’t mean to diminish it, but it’s not necessarily core enterprise resource planning (ERP), if you will.
SM: Why do you think that’s the case? Is it because core ERP has a lot of exit barriers to integration because it already happened, it is already entrenched in the organization? Is there no real business driver for moving core ERP into the cloud? Is that the right assessment?
MW: To answer that, I need to divide it into quadrants. One dividing line would be small and medium businesses, or perhaps more significantly, startup businesses from large enterprises.
SM: And ERP for small businesses is going to the cloud . . .
MW: Exactly, and even small businesses. Look at how venture capitalists and private equity firms are funding businesses. They are giving very little capital for things such as IT. On the enterprise side, adoption is less toward enterprise applications and more toward non-core or edge applications. That would be the horizontal dividing line. The vertical dividing line would be the public cloud versus the private cloud. So, the comment I made was more about the public cloud. If I look at the private cloud, adoption of large-scale apps is certainly underway because of two factors. One, it is inside the trust zone. And that’s frankly the more subtle definition of a public–private hybrid community that makes more sense. It’s those workloads that occur inside the trust zone.
SM: So, let me probe into the large enterprise private cloud. What workloads are moving?
MW: Capabilities or workloads that look like storage in the cloud, because that’s a clean extension of the storage area networking technologies that we have been doing. Driving that, frankly, is the analytics cloud, the messaging cloud, and the collaboration cloud. Those two are slightly different, but for the purposes of this conversation, we will join them. And then, the classic enterprise application clouds, which are CRM, sales field automation, and customer care, which for many companies is at the edge. And, cloud infrastructure increasingly being used as a service to run core ERP. So, to summarize, [the movement] is in private cloud large-scale enterprise, infrastructure as a service, and platform as a service not necessarily software as a service. Big ERPs are running on cloud infrastructure privately.