Someone sent me a link to this article this morning. While reading it I started to think about the competitive landscape for cloud service providers and the challenges telcos have competing in this market. After all, when you think cloud service provider, which names come to mind? Amazon? Microsoft? IBM? Google? Sure, but how about CenturyLink? AT&T? Verizon? What are the key differentiators between the first and second groupings? The first group is closely associated with software and innovation, while the second is associated with dropped calls, poor customer service experiences and expensive telecommunications services.
Understandably, telcos have a perception challenge being considered a competitor on par with technologically innovative firms. However, these vendors have a strong value proposition being linked to the core telecommunications infrastructure (can we say cloud doesn’t work without network?) and deep experience operating hosted infrastructure on behalf of business for the better part of the last three decades. Indeed, many businesses still have a significant base of compute infrastructure being operated by telcos today.
One name that I particularly see as a standout here is CenturyLink. Their acquisitions of Savvis, AppFog and Tier3 have provided them with a solid foundation for delivery of cloud computing services including both Infrastructure- and Platform-as-a-Service offerings. CenturyLink’s approach to cloud differs a bit from the commodity public cloud service provider approach in that they focusing on an enterprise user that is looking for an integrated offering versus the AWS and Azure users that seem to appreciate a do-it-yourself approach to building an operational environment for applications. Hence, it is difficult to do an apples-to-apples comparison solely based on price as compute along will be three times as much as AWS, but includes capabilities not included by default in AWS, such as monitoring, dedicated hardware and support. Of note, this model will see growing pressure now that IBM is offering up dedicated hardware on demand through their SoftLayer division.
This approach, let’s call it external private cloud, seems to be quite popular among operational staff as a common ground for obtaining the benefits of enterprise on-demand or OpEx computing and the traditional operational controls enterprise IT operational folks are familiar and comfortable with using. I find this very interesting when you consider what’s happening in with the rise in interest of DevOps and the significant base of developers flocking to the pure public cloud service provider experience, such as that offered by AWS.
While I don’t see an issue with leveraging on-demand public cloud for development and testing and external private cloud for production, the tools to make this seamless are only now starting to become available and are highly immature for this type of activity. Moreover, it continues to model the internal development/production differences that we see today among businesses still managing their own internal infrastructure. Thus, the Dev/Ops partition continues unabated leading to common problems of building for one environment and deploying to production in another.
But I digress… Back to the point at hand.
The offerings from like likes of Amazon, Microsoft, IBM and Google will continue to grow in use as these are often the first place users go when they are looking for on-demand compute services. That said, when it comes time for enterprises to deploy production versions of applications they deem operationally important, it shouldn’t be a surprise to find that they are probably talking to their telco provider. Of note, realize that the telco provider also has favorable sourcing status with procurement given that they are providing networking and telecommunications services. Getting Amazon or Google added to those lists might seem like no-brainers, but typically, sourcing is not seduced by sexy tech vendor names.