X
96843

Hyperscaler as a service: service providers shift from competing to partnering

April 10 2019
by Al Sadowski


Introduction


Over the last decade, the relationship between the major public cloud providers (AWS, Microsoft Azure, Google Compute and IBM, among others – aka the hyperscalers) and the next tier of public cloud providers (systems integrators, telcos, MSPs and asset-heavy SaaS providers) has evolved. These relationships have shifted from competition to partnerships – hyperscaler as a service. This evolution represents recognition by these service providers that trying to go head-to-head with the hyperscalers can be a fool's errand.

The 451 Take


The hyperscalers have economies of scale when it comes to purchasing hardware and expanding their datacenter footprints geographically. They have armies of engineers cranking out new products and features at an increasing cadence. With their growing ranks of security, application and vertical specialists, enterprises are finding more use cases that are being evaluated for migration to these public cloud venues. Trying to compete on price or trying to differentiate solely on geographic footprint has proved to be a losing strategy for other service providers, which are now shifting their IT budgets, staffing and product roadmaps so that major public cloud providers' functionalities and footprints are part of their own products and customer-facing solutions.

Context


According to 451 Research's Differentiation & Vendor Selection 2018, 86% of telcos, MSPs, systems integrators, MSPs, colos and SaaS providers are currently partnering with the largest of public cloud providers. This is a major shift from 7-10 years ago, when service providers were in a race to provide compute- and storage-based services for the lowest price. Without the buying power of the hyperscalers, and having such headwinds as legacy support systems, operations and sales focus, many providers that attempted to compete with AWS, Microsoft Azure and Google on price have since exited the market. Over the last three to five years, service providers tried to differentiate with their regional footprints, data sovereignty expertise, and other compliance-based or application expertise. As the hyperscalers spent billions on geographic expansion and the development of more products and services, the value propositions for these tier two and three service providers eventually became obsolete. These service providers have turned to a strategy that embraces the hyperscalers as a partner and for the services behind the service.

Figure 1
Differentiation & Vendor Selection 2018

Hyperscaler under the hood


The largest public cloud providers are churning out new products and services at a breakneck pace. Their smaller competitors are unable to keep up, so they have changed tactics by building products and services that incorporate the hyperscalers' own capabilities. For example, rather than trying to duplicate the functionality, 38% of service providers are integrating AWS, Microsoft Azure, Google and IBM's machine learning and artificial intelligence functions into their own offerings. This is being done to offer ML as a service and related service catalogs, in addition to incorporating AI/ML functions into their customer-care and operational use cases. This same premise applies to functions as a service (aka serverless). Rather than build their own, 36% of service providers are using AWS Lambda functions and related services from other hyperscalers under the hood of their own customer-facing applications.

Hyperscaler hybrid solutions


451 Research's Cloud, Hosting & Managed Services, Vendor Evaluations 2018 indicates that 55% of enterprises globally are moving toward a hybrid IT environment that leverages both on-premises systems and off-premises cloud/hosted resources in an integrated fashion. The percentage is even higher for larger enterprises.

Besides the examples above about incorporating hyperscaler capabilities into their own applications, the once-competing service providers have accepted that the largest public cloud providers are a preferred destination for some enterprise workloads and data. As such, 41% of service providers have established network connectivity from their own datacenters to the hyperscaler datacenters. Many are also building and buying the software to offer hybrid cloud and multi-cloud management. By offering enterprises a one-stop shop to orchestrate and manage compute and storage resources for on-premises, hosted services, as well as network access to a variety of public cloud venues, service providers remain relevant for enterprise digital transformation opportunities.

Managed hyperscaler


Rackspace (with its acquisition of Datapipe) recognized an opportunity to provide enterprises with the management services for AWS, Microsoft Azure and Google Cloud, rather than trying to compete head-to-head for IaaS. Many other service providers have adopted a similar strategy. According to our research, 41% of service providers are also offering managed public cloud services.

While it is one thing to offer the managed services, it is another to actually make the necessary investment to succeed. As such, 42% of all service providers are planning to increase headcount for third-party cloud management, while only 11% are expecting a decrease in these teams over the next 12 months. These moves allow enterprises to focus on their core business without having to train limited staff on the various hyperscaler management consoles and pricing options.

Outlook


We expect that more service providers will incorporate additional hyperscaler services into their own products – for example, security functions, container services and data analytics. Additionally, IBM, Oracle and Alibaba will be partnership targets for the MSPs, telcos, SIs and SaaS providers looking to diversify their portfolios on a global basis and for specialized workloads – more hyperscalers as a service.