Cloud

A New Cloud Price War Unfolds

The next wave of cloud cost reductions will be driven by targeted partnerships and investments. Explore the impact to the cloud ecosystem and customers with IDC's Deepak Mohan.
Pinterest LinkedIn Tumblr

Value proposition of cloud expands over time, from ‘cost efficiency’ to ‘new services and technologies’

The first decade of cloud infrastructure services, from the mid-2000s to the early 2010s, was marked by an intense focus on the ‘cost efficiency’ value proposition of cloud. Cost savings to customers may be delivered in different forms, such as elimination of upfront costs or lower overall cost, but this was arguably seen as the top benefit of moving to the cloud. This was also reflected in the initial target customers to whom cloud appealed, typically shadow IT developers, startups, or other segments (like hobbyists) that prioritized cost efficiency.

This attention to cost is reflected in the announcements and focus topics of the period. Price reductions and lower cost instance types, for example, were an important and an eagerly anticipated part of Amazon Web Services (AWS) announcements. These were often highlighted in communications, emphasizing the number and magnitude of price reductions as an indicator of value delivered to customers. This focus on cost also influenced competitive actions in the market, marked by a series of “cloud price wars” through the early 2010s.

Today, the cloud value proposition goes well beyond cost optimization. IDC’s IaaSView 2019 reveals that less than half of the respondents used total cost of ownership as a top factor in evaluating cloud providers. Other important value propositions prioritized by customers include access to new technologies, speed of scaling, and geographic reach. This shift is also reflected in AWS announcements. Between 2009 and 2018 the proportion of AWS “what’s new” updates that were related to pricing updates dropped from approximately 10% to under 3%, reflecting the steadily expanding focus on new technologies and higher layer services like AI, edge, and robotics.

Cost efficiency starts to resurge in priority, but traditional cost optimization levers are saturating

Cost efficiency, however, continues to be important characteristic of cloud services. It had just temporarily transitioned into a table stakes expectation from customers. However, the focus on cost efficiency can be expected to increase in priority as the market ramps into its next phase of growth in cloud usage. This phase will be led by enterprise IT workloads being deployed in cloud by the median enterprise, and the rapid movement of enterprise IT workloads into cloud with varying levels of modernization.

As the proportion of the IT budget allocated to cloud increases, the focus on cloud costs will increase, and cost efficiency will once again rise as an area of focus and a decision driver for customers. And it is reasonable to expect this focus on costs to drive yet another wave of price reductions and price wars among cloud providers.

But cloud providers have challenges to overcome, to deliver the next wave of cost savings. In the early growth years, incremental cost optimizations were enabled by a mix of Moore’s law, architectural optimizations, and economies of scale. These cost optimization levers, however, are nearing saturation in terms of the cost benefits they can deliver.

With Moore’s law, while the directional trend continues to hold, in terms of continuing reductions in transistor sizes and increasing chipset densities, its pace has slowed. With fabrication technologies already starting to go below the 10nm process, this starts approaching the physical limits possible with transistor size reductions. Another major lever for cost structure improvement has been the architectural and supply chain optimizations possible at large scale, such as using internally developed software and systems to deliver the cloud service, resulting in a high level of ability to customize and optimize the stack for the providers operations. Most major cloud providers now have substantial investments in the underlying software and hardware systems used to deliver their services, and this lever too is close to saturation.

Another related factor that helped was the growing economies of scale improvement during the initial years. As the customer base and usage grew, from hundreds to thousands to millions to tens of millions of customers, providers were able to unlock value faster from investments and shared resources – both in terms of mega datacenter operations and in terms of overall delivery of services. But the operational efficiency gains slow down with scale, and the economies of scale benefits at providers have also started to saturate. Moreover, newer service delivery models – particularly those delivered outside the providers mega datacenters – may even have the effect of increasing the average unit delivery cost of services for providers.

New levers emerge to power the next wave of cloud cost efficiencies

With the saturation of these levers, providers have since looked for new approaches to drive cost structure improvements and to enable cost savings on services. IDC has observed three broad approaches being used by providers, as they look to enable the next chapter of cost structure improvements. While there may be overlaps across these approaches, each represents a distinctly different way in which the value is delivered to customers. These are discussed below, in the increasing order of impact they can bring to the overall cloud market.

  1. Improvement of the overall cost of operations through software partnerships, integrations, and bundles. From a customer perspective, the overall cost of operations for a workload includes both the cloud services cost as well as the development, licensing and maintenance costs for the application software. IDC research on workloads highlights that nearly half of all the workload being deployed on public cloud are commercial off the shelf software packages. For such software, the cost of integration, licensing and maintenance often exceed the underlying infrastructure costs, representing the majority of the overall cost of operations for the workload.

    Cloud providers can address this overall cost of ownership with pre-qualified, integrated or bundled solutions optimized for specific software workloads, including license mobility options and price-performance benefits, allowing customers to benefit from improved cost of ownership and cost of operations for the workload.
  2. Focus on targeted and use-case specific optimizations. Cloud usage has expanded to nearly all types of IT use cases. The increasing scale of usage now allows cloud providers to identify specific use cases with sufficient critical mass and invest in targeted optimization for these specific use cases. These may include identifying a niche segment where the provider can deliver a differentiated solution, or targeted pricing changes to optimize for certain types of use cases.
  3. Continue to improve on economies of scale benefits by going deeper into the inbound supply chain for the cloud service. While in-house development of infrastructure systems and software enables cost and performance optimizations, these still rely quite heavily on underlying components from other vendors, including processors and coprocessors, memory, and storage media.

    Key among these are processor and co-processor technologies, both of which are markets dominated by a single technology vendor each. Reducing reliance on this dominance would allow higher control on cost structure. While this requires the highest level of investment among the three discussed, this also promises the broadest level of impact in terms of delivering cost efficiencies.

IDC believes the next era of cost structure improvements and price reductions will be delivered through these mechanisms. As mentioned earlier, these three do not necessarily represent strict demarcations in the types of investments by the providers. But these do represent differences in how the cost savings are delivered to the market and in how customers can benefit from these cost efficiencies.

Implications for customers and technology providers

The targeted nature of these cost efficiencies makes the pace of availability of these cost savings highly use case and customer workload specific. Unlike the cloud price wars in the 2010s, these cost efficiency investments will not automatically translate into benefits for the entire customer base. While it is true that these changes and cost improvements will be gradual in their emergence and will take years to become generally applicable, it is important to remember application lifecycles are typically even longer. Customers must therefore pay attention to application dependency choices being made today, as well as software and cloud provider partner selection, to maintain the broadest set of options for their specific workloads and use cases.

For third party technology providers, particularly those that want to increase participation in the cloud ecosystem, these new directions of investments create new opportunities in terms of collaboration and partnership with cloud providers. The use case specific focus and the attempts to improve overall cost of operations for the customer will increase the willingness from cloud providers to invest in deeper engagements with third parties, as they attempt to address and deliver cost savings to an ever expanding set of fragmented customer use cases and workloads.

These factors don’t take into account the effect of COVID-19 on the cloud infrastructure market. See how the related economic disruptions will affect the cloud market; read “Worldwide Cloud System and Service Management Software Forecast, 2020–2024: COVID-19 Disrupts Growth Trends”:

Deepak Mohan

Research Director, Cloud Infrastructure Services