Is the cheapest cloud pricing flexible or granular?

As maturation of cloud computing occurs, IT is looking for pricing that allows them to maintain value as their needs change.

cloudvson premerp
istock

Cloud computing has quickly expanded as a core IT cost over the course of this decade as the ease of implementation, scalability, geographic availability and improved security have allowed enterprises to use cloud for net-new computing uses, expansions of existing compute use cases, and even replacements of on-premises data centers in some cases.

As the use of cloud computing has basically doubled in the past couple of years and will double again by the end of 2020, the pricing of cloud computing resources has traditionally been granular in nature based on per gigabyte or per minute billing of resources. Recently, Amazon has added to this granularity by providing per-second billing for a variety of services, a billing change that Google Cloud was quick to imitate. This billing model is reminiscent of the traditional mainframe measurement of MIPS (Million Instructions Per Second) used to measure the cost of computing. Because IT is naturally focused on transactions, disk space and processing cycles, it is apparent why cloud computing billing would focus on these granular units. This mindset reflects the fact that the cloud buyer has traditionally been an enterprise architect, project manager or line-of-business based techie focused on finding a set of cloud-based services to support a specific project or initiative.

This pricing model is very useful for small projects, as the incremental cost of short-term computing workloads is much cheaper on the cloud. And this model also works well for predictable peak-periods, such as holiday seasons (Singles’ Day, Black Friday, Christmas) where peak usage is significantly higher than standard operational demands. However, it also provides some challenges from a financial perspective.

Those who are familiar with telecom expense management know how the challenges of managing network traffic, data plans, voice minutes and the day-to-day management of assets assigned to specific employees, locations, departments and projects can be both a mind-numbingly challenging management task and an opportunity to save millions of dollars by shutting down overprovisioned locations and zero-usage accounts while optimizing existing services. Perhaps the simplest way to think about this from an expense and accounting perspective is that data center and server purchases are being replaced with a telecom expense model. The increased accounting challenges will be paired up with the ability to drive additional savings.

As cloud starts to shift from a departmental enabler to a foundational enterprise technology and enterprises start seeing their cloud computing bills shift from the thousands of dollars that fit into an expense to a multi-million dollar budget item, the buying mode for cloud will shift as well. As these contracts get renewed, cloud will start to shift from a service that can be bought and discontinued at will to an enterprise contract that must fit within corporate procurement, IT-based service level agreements and a strategic technology roadmap.

As this maturation of cloud computing occurs, IT starts looking for pricing that allows them to maintain value as their needs change. All this leads to what I believe is the most interesting change in cloud pricing to this date: Oracle’s Universal Credits that will allow for usage “credits” to be transferred between Oracle’s Infrastructure as a Service (IaaS) and Platform as a Service (PaaS) products. This flexibility serves several purposes.

  • Computing flexibility: First, these Universal Credits allow IT organizations to abstract their infrastructure and platform spend, which makes sense considering that there is a continuously changing spectrum of distributed vs. centralized computing. Just as we have swung from mainframe computing to client-server computing and now back to a mobile-cloud computing paradigm, there will eventually be a time when edge computing and storage become more important again. (This change will be driven by the Internet of Things and blockchain and is close at hand, but I think we have another 12 to 18 months before the change is really noticed in earnest by mainstream enterprises.)
  • Financial predictability: Second, these Universal Credits help provide more predictable IT spend. In the business world, predictability is worth a certain premium over unpredictable a la carte spending. Even if the latter model ends up being cheaper, the inability to predict and budget for these costs leads to the inability to plan and to accurately allocate budget to other technical and operational needs.
  • Staffing flexibility: Third, this infrastructure flexibility also helps companies that are using contingent labor and consultants in their infrastructure environments. It is easier to align talent resources and services based on current business needs. Although “the cloud” removes the need for direct server and resource management, there are still related data management, analytics, developer, end user computing and project management resources that can be scaled up or down depending on the current quantity and variety of IaaS and PaaS accounts currently provisioned.
  • Contractual flexibility: Although Oracle’s Cloud is relatively new compared to current market leaders Amazon, Microsoft, Google and IBM, Oracle’s ability to support IaaS and PaaS at scale is a distinct advantage. If used as a single provider for both IaaS and PaaS, IT procurement has the ability to negotiate consistent and potentially interconnected SLAs across both sets of services to gain a higher level of support than could be expected from separate IaaS and PaaS vendors. Because the same credits are being used for both IaaS and SaaS, the client’s IT expectations have to be supported in a consistent fashion over both services.

All this being said, we are talking about a pricing and contracting change that can be imitated by other vendors. I would expect that Amazon Web Services and Google Cloud Platform will continue to work on a granular basis in charging by the second and perhaps digging down into milliseconds, megabytes or whatever unit provides additional specificity. Based on my two decades of experience on telecom and IT billing across both providers and enterprises, I think that rate cuts and increasing pricing granularity will continue. However, these cuts will reflect the increasing efficiencies of scale that these vendors build out and will not provide massive benefits. This pricing is built on attracting initial business and earning lock-in. But there is a reason that Amazon Web Services is currently the main source of operating profit for Amazon as a while: cloud computing is very profitable as it is currently priced, bought and used. Amalgam Insights estimates that in the average enterprise cloud environment that has been up for 2+ years, 30 percent of current services is either unused or represents excess and unnecessary capacity. Not surprisingly, this matches up well with the around 30 percent margin that Amazon receives on Amazon Web Services.

Oracle is leading the way as the first cloud provider to provide true enterprise pricing options. I believe Microsoft should follow suit, as it can compete in some respects and Azure clients need the flexibility to build more robust and complicated applications on the Azure Cloud over time. It would make sense for Microsoft to support the need for flexibility for established clients. IBM should also take this route with the additional ability to expand credits to other products such as the zSeries mainframes.

This split would lead to the first true vendor bifurcation in cloud infrastructure as Amazon Web Services and Google Cloud focus on the technical developer and architect buyers and Oracle, Microsoft and IBM focus on the business buyers who seek to take advantage of purchasing at scale to gain consistency and broad-based support. There is room for both approaches as the cloud computing market starts to engulf the traditional data center market. Whether this happens or not, I’ve provided my perspective.

Regardless, Oracle has provided a real alternative to its competitors in offering a Universal Credit approach across IaaS and PaaS. Although it’s not going to be an apples-to-apples comparison, I think that fast moving companies that understand their organization’s current positioning on the “edge vs. distributed” computing spectrum will be well suited to work with Oracle and provide a broad cloud services toolkit. Although I would recommend for other cloud infrastructure providers to follow suit, I believe that the complexity of supporting this model will be difficult to quickly replicate in the near term.

As much as we all talk about the cloud, it is easy to forget that the cloud computing market is still relatively new. As providers start to figure out exactly what their business models should truly be to provide consumer satisfaction and success, my key recommendation to fiscally responsible IT departments is to follow the pricing units. Pricing is a key differentiator in cloud, but the differences are less about understanding 2.63 vs. 2.64 pricing differences and much more about tracking the units, transferability, and service levels across all relevant vendors being considered.

In conclusion, my key recommendation for CIOs trying to get cloud expenses under control is: Look at how your cloud services are being charged and decide whether the pricing is aligned with your preferred model of cloud procurement and management. This business decision will differentiate executive CIOs from operational CIOs in managing the cloud. CIO-based cloud management decisions will end up fundamentally shaping the strategic importance of IT managers. Choose wisely.

This article is published as part of the IDG Contributor Network. Want to Join?

Participate in the 2019 State of the CIO survey. Make your voice heard.