Private clouds are forcing IT to question who pays for what.
Organizations moving their IT infrastructure to private clouds are encountering demands from end users to pay only for the resources they use -- as cloud vendors suggest. This is affecting IT funding and forcing IT to adopt new chargeback methods.
Traditionally, IT has split the amortized costs of its infrastructure across all groups within the company that use its resources. In a private cloud network, how do you do this so you can make sure that IT stays viable? It’s a problem confronting many CIOs.
For years, IT simply amortized datacenter floor space, software, and equipment costs over time -- and added to this the charges of salaries and benefits for staff to come up with an annual dollar figure for funding IT. Funding came out of corporate revenue, cash on hand, or reserves, depending on the enterprise’s financial picture.
Later, accounting systems got fancier so that these IT costs could be assigned to the lines of business that brought in the revenue. However, a fair accounting of IT resources remained elusive. In one case, I remember a CIO telling me that his company booked all of its IT costs to its mainframe, because it was the only system that had cost tracking on it! Consequently, all of the non-mainframe resources the company was using appeared to be “free,” while the mainframe seemed to be very expensive.
“Sticks and glue” solutions like these won’t work with private clouds. Business users are going to push for “pay-per-use” chargebacks for IT services, thanks to competing vendors in the open marketplace that adhere to this model.
Here are some steps can IT take to adjust its funding mechanisms to fit today’s pay-per-use trend:
1. Revisit the tracking of IT expenses. Most datacenters still track IT resources as “fixed” entities -- whether these resources are pieces of hardware, application software packages, or the software needed to operate and manage IT infrastructure. Today, however, new IT infrastructure management software can automatically track workloads across all of these resources and arrive at more accurate cost figures for what a given line of business or enterprise area is using. For example, if Accounting wants you to run its general ledger, payables, and receivables, infrastructure software with automation can analyze all of these applications and track the consumption of IT resources (e.g., CPU, storage, networks, etc.) required to run these systems. The infrastructure software can also consolidate all resource usage during a given period as a “per use” chargeback to Accounting. Of course, IT has to set up the process by assigning a dollar value to each consumable resource so it can be calculated into the workload.
2. Identify fixed costs. Every business has fixed costs, and business end users know this. They also know that a commercial cloud services provider is going to factor in some of its fixed costs into pricing. Likewise, it is fair for IT to factor in the costs of its support staff and facilities. Fixed costs can be amortized and computed proportionally into the monthly pay-per-use charges for IT services./
3. Isolate the “extra” service items and let business areas pay for them à la carte. When the organization goes to a cloud services provider and asks for additional training and implementation, the items are billed separately. IT should have a similar model that adds these costs onto its baseline pay-per use-fees.
4. Factor in some profit. Even if you are an internal IT organization, it is smart to set aside some extra dollars for IT prototype and R&D work that can bring technology advantages to the business. Commercial technology companies typically reserve 15 percent of revenues for technology reinvestment. Best-in-class IT executives also find ways to fund some R&D, because it can deliver a competitive edge to the company. However, to do this, the CIO must first achieve buy-in with key enterprise executives on how much to “upcharge” services to fund the R&D.
— Mary E. Shacklett, President, Transworld Data