Never Miss a Beat: Get a snapshot of the issues affecting the IT industry straight to your inbox.
The Financial Future of Data CentersThe Financial Future of Data Centers
Users want to see where they get more bang for their IT buck. Can we use pay-per-use costing for internal data centers?
5 Min Read
JL via Alamy Stock
At a time when companies are moving IT to the cloud, and when multi-cloud IT deployments are growing, companies are keeping their on-premises data centers.
I’ve queried CIOs on why on-prem data centers are surviving, and these are the reasons that they give:
Internal control over mission-critical applications;
Cheaper processing costs and faster transaction times internally instead of by routing transactions to and from the cloud over expensive telecommunications links;
Better security;
Internal integration complexities that would make it difficult to migrate systems or ecosystems of systems to the cloud.
All those CIOs also agree that the cloud is a compelling solution because of its real-time ability to downscale or upscale, and its value proposition of charging you only for the resources and services that you consume.
The end result is that on-premises computing and cloud computing are co-existing, although admittedly, the persistence of internal data centers contrasts with the futuristic vision of “all-cloud” IT operations that many IT practitioners foresee.
The present computing vision is constructed around hybrid computing, which enables sites to use a variety of both on-premises and cloud-based IT. In some cases, applications will span multiple clouds and on-premises resources with the processing they do. In this world, CFOs and IT leaders are beginning to think about the dollars and cents side of how IT value is delivered to internal user-customers, and how it should be budgeted and charged back.
Users now want to pay for only what they use. They, and many upper-level executives, are no longer content to see IT as a “sunk” cost that gets baked into the overall enterprise budget where everyone ends up paying for not only for what they use, but also what others are using. Some users believe that they are paying for more than they should.
The solution?
Reimagine how on-premises data center costs are calculated, and then convert them into more of a pay-per-use model along the lines of what users are seeing on the cloud.
Three new trends are facilitating this change of thought:
More IT departments are transforming on-premises IT into a private cloud that they operate and maintain, and the private cloud operates on the same pay-per-use premise as an outside cloud would;
More IT vendors are allowing sites to fund their internal data center hardware and software on a monthly rental basis that can be expensed operationally instead of amortized and depreciated over time as with purchased assets. Under these rental plans, equipment can be swapped out and upgraded, eliminating the issue of wasting data center assets that lose value on the books through depreciation and obsolescence, and that can’t do the job anymore;
More vendors are offering flexibility in IT deployments. For instance, if you’re running a system on internal hardware and you want to move it to the cloud, you can. Conversely, you can move off the cloud and back to on-premises computing. In both cases, a growing field of vendors is offering pay-per-use plans that operate in both on-cloud and on-premises environments.
Financially, what companies end up with is a kind of “hybrid costing” model that goes along with hybrid computing.
Are companies there yet?
No, but more are warming to the idea of pay-per-use costing, whether it comes from the use of internal data center resources or an external cloud.
A cloud provider already allocates costs based upon the number of resources and staff hours, facility space, power, etc., that are consumed by a given user group. If a similar pay-per-use cost can be calculated for the on-premises data center, with monthly data center and outside cloud costs totaled for each user department, everyone gets better cost visibility, and users are more likely to feel that they are getting the value they’re paying for.
This monthly user department charging concept isn’t exactly a new idea. It was first introduced by computer service bureaus in the 1960s, with cost allocation formulas that assigned a percentage (and amount) of monthly processing cost to each user area. In most companies, IT then moved away from this model, instead classifying IT as an internal enterprise cost center that the enterprise as a whole funded. The old data center chargeback idea is now reinventing itself as a pay-peruse concept that closely aligns with the cost models that external cloud providers use; and users are receptive to it.
On the IT side, this provides significant flexibility if, for example, IT chooses to co-locate servers and storage in a third-party data center instead of running these assets in its enterprise data center. Conversely IT assets (and applications) can be moved back to the corporate data center if there is a call for it. What happens from an IT perspective is that the word “data center,” formerly defined as the internal physical computing center of an enterprise, expands beyond the physical boundaries of the enterprise and IT begins to think more broadly about the definition of a data center.
Vendors see this reinvention, too. In January 2024, IBM said, “The evolution of the data center helps position your organization at the forefront of technological advancement and at the heart of sustainable business practices.” In this environment, a clear implication is that a hybrid combination of cloud-based and internal data center IT will collectively redefine what a data center is, and with that IT transformation, new financial costing formulas will emerge.
About the Author
You May Also Like
More Insights
