We're starting to see a lot of change in the infrastructure outsourcing space, much of it as components of cloud computing. We spoke with three service providers in the IT infrastructure space -- CompuCom, HP, and Satyam -- to learn their insights on where today's changes are heading as far as their influence on service delivery and value proposition in the next few years.
Satyam, a global provider based in India, notes that remote infrastructure management outsourcing (RIMO) is opening up to delivery from offshore providers.
The North American economic environment is a significant driver of the acceleration of the RIMO model, and the cost benefit of labor arbitrage with offshore providers is attractive. "A lot of companies are now willing to change their model for managing infrastructure," says Anirban Dutta, head of strategic deals for high tech verticals at Satyam. "This is especially true in the mid-market, which is now focusing on RIMO as a key part of their business transformation initiatives."
With demand increasing, there is global play and aggressive competition among IT providers. That's good news for buyers, notes John McKenna, chief services and strategy officer at CompuCom. He predicts a lot of competition in the next three years among providers, which will influence the cost and quality of infrastructure solutions.
"Today you can't do BPO without ITO. As the two converge, tomorrow you won't be able to do ITO without BPO," states McKenna. And that creates a challenge. Companies have understood and documented business processes since the early ERP days 20 years ago, but infrastructure processes are less understood and documented.
The industry is now beginning to approach infrastructure from a stronger process orientation. ITIL and ISO 20000 are driving much more structure in infrastructure management. "As it becomes more process centric and the processes are more consistently defined, the services can morph," says McKenna. With consistent definitions and clarity around the inputs and outputs to a help desk, for example, service providers will be able to plug and play different kinds of services. The consistent, process-centric structure will then enable clients to have more agility. "They'll be able to drive more innovation and select different providers in different areas on a more frequent basis," he says.
He recalls, "Over the years we've seen a plethora of examples where a process wasn't clearly defined before it was offshored, or a process wasn't defined before it was automated, or a process wasn't defined before someone attempted to make it plug and play. Those efforts were not successful."
As managing infrastructure as a process evolves over the next five years, McKenna says we'll see "a lot of infrastructure process reengineering to deliver optimized performance." He explains that the next step beyond labor arbitrage is the automation of infrastructure processes. That automation can only happen successfully if there is a process. He likens infrastructure process reengineering to the BPR and ERP process optimization efforts of 20 years ago and predicts we'll see the same type of industry and tools grow up around infrastructure processes.
But there are two differences between the infrastructure process reengineering and the BPR/ERP of two decades ago. Unlike SAP's leadership in ERP, "as of now, there is no SAP equivalent in infrastructure," says McKenna. Secondly, an infrastructure delivery platform will be necessary to enable rapid and effective infrastructure process reengineering.
As an extension of that platform, we'll also see avatars for service desk agents to assess infrastructure, McKenna predicts. The avatars will be enabled by ancillary technologies bolted onto the delivery platform and will perform such tasks as remotely resolving basic infrastructure issues and automating and resetting passwords.
Managing infrastructure as a process is not, for example, giving a whole service desk to one provider to manage. When managed as a process, the buyer can swap out the service provider every day if it wants to. "Managing infrastructure as a process is looking at how logical connections of infrastructure processes and systems link together," explains McKenna. "For example, in end-user computing (not data center computing), you might have a process that starts with the end user, goes to the service desk, goes to whatever resolution process it needs, and gets back to the end user as one integrated process from back to front.
Price and cost models will change. McKenna predicts that as infrastructure management becomes more process centric next year, cost and price models will move from a per-seat to a per-service price.
Advantages for buyers include lower risk, consistent service, agility, flexibility to change service providers, and a greatly enhanced capability for innovation. But it can't happen without the integrated platform -- which requires clearly defined infrastructure processes. McKenna describes the current state of defining the processes as "still in very rudimentary stages."
Although the industry has been talking about Infrastructure-as-a-Service (IaaS) for a couple of years, many companies viewed it as something still on the frontier. It's a reality today, although it's still in the early adopter stage. HP's offering in this arena is "Adaptive Infrastructure as a Service."
New outsourcing delivery models that provide a high level of flexibility for on-demand services and enable clients to avoid extending capital and "the pain of building a data center and managing a computing environment has always had great appeal," says Bob Pryor, senior vice president, Sales and Marketing, HP Outsourcing Services. "The challenge is putting it together with enough robustness, discipline, security, and predictability for an enterprise client, especially for anything approaching an important business application, much less a mission-critical application."
HP's Adaptive Infrastructure service (currently available in Europe and North America) is a preconfigured, shared, secure, high-availability environment that outsourcing clients can access for long- and short-term capacity for their business applications. It's available in a subscription pricing model, making it highly flexible. "It's not a vanilla environment," says Pryor. "It's a highly standardized, highly configured environment that can run mission-critical SAP or Microsoft Exchange applications for clients. It can also support clients' other applications with optional management services." HP recently signed its first large client for the Adaptive Infrastructure service.
Start the IaaS conversation now
Adoption of IaaS services is growing at CompuCom. Infrastructure conversations with clients often start with a request to upgrade one piece of infrastructure management software. "We point out that they can spend the same or perhaps just a little more for an IaaS model and get much more than one piece of their infrastructure under management," says McKenna.
He adds that if clients want to get to that model even five years from now, it's not too early to start the conversations today.
Meanwhile, service providers are investing in automation for infrastructure management. CompuCom, for example, is looking at periphery technologies that will enable a better end-user experience in a self-service portal. "Avatars are a technology that drives adoption of self-service," says McKenna. "Self-service remains the most difficult aspect to make happen in infrastructure management because you have to change the behavior of the end users." He says the company is also investing in real-time service-quality metrics to increase transparency for clients.
What comes next after infrastructure-as-a-service?
"Today pricing is per seat, tomorrow it will be per service, and in the next year I believe we'll see a move toward a per-information price," states McKenna. He predicts that infrastructure management will become information management and over the next two years client discussions will start to revolve around information-as-a-service. "In that model, you'll be able to plug and play according to the information technology services you want to deliver to your end users," he says. "The model will provide the fastest time to market for the latest infrastructure processes. It will be the ultimate in innovation and flexibility."
Lessons from the Outsourcing Journal:
- ITIL and ISO 20000 are driving the outsourcing industry to begin to approach infrastructure from a stronger process orientation. This will result in consistent definitions and clarity around the inputs and outputs to a process, enabling service providers with more agility for plug-and-play services, lower risk, more consistent service, and enhanced capability for innovation. It will also enable buyers to quickly and easily change providers.
- Look for infrastructure process reengineering and development of management tools over the next few years (much like business process reengineering 20 years ago) in an effort to deliver optimized performance.
- As infrastructure management becomes more process centric, price models will move from a per-seat to a per-service price.
- Infrastructure-as-a-service (IaaS) is a reality today, although it's still in the early adopter phase. Look for more providers to develop on-demand offerings for IaaS.
- Self-service remains the most difficult aspect of infrastructure management because it requires changing end-user behavior. Some providers are investing in avatars to help drive adoption of self-service.