tashatuvango - Fotolia
At its recent re:Invent conference in Las Vegas, Amazon unveiled that Turner Broadcasting System chose AWS as its "preferred cloud provider." Turner, which controls networks like TBS, TNT and Cartoon Network, said it would migrate decades' worth of content to the AWS cloud, including CNN's 15 petabyte video archive, one of the biggest in the world.
Turner also plans to move thousands of VMs to AWS, and it intends to use a wide range of Amazon AI technologies to better analyze and extract video metadata. Turner hopes to provide better personalization to its viewers, as well as better real-time analysis to advertisers.
Turner is the latest addition to Amazon's customer list -- which now also includes Hess, Capital One, Kellogg's, Siemens, Adobe and Comcast -- to suggest that AWS' cloud strategy attracts not only startups, but enterprise buyers as well. It makes sense for these Global 2000 companies to use AWS, considering its lofty position in the cloud market, and I suspect AWS has locked up many other large companies but is unwilling to discuss details.
A big part of AWS' cloud strategy is to lock up large companies to push even further into the enterprise market. Google, Microsoft, IBM and other cloud providers play this game as well, but AWS seems to be picking up the pace. And when you consider that the other big public cloud players were enterprise software players before AWS existed, it displays how much of a fantastical feat Amazon has pulled off.
Keep your options open
Should your enterprise execute this kind of preferred cloud provider deal? I suspect that a discount comes with these agreements, which is one clear benefit. These agreements might also generate some publicity, a big factor in AWS' cloud strategy. But nobody should sign an exclusive deal. There should always be the option to use other public clouds in the future.
In some cases, AWS' cloud strategy might not be a match for the technology and business needs of a particular customer, such those that only need infrastructure essentials, like storage and compute, and not the higher-level platform services AWS has focused on recently. Cheaper upstarts, like Alibaba, might find their market among these infrastructure-focused clients. These enterprises might also be better served with a managed service provider or even an on-premises option. Companies usually strike these kinds of preferred vendor deals within boardrooms, not IT departments, which work from a specific set of requirements. This can result in knee-jerk decisions that might not be best for a project's needs.
That said, companies that select AWS or another public cloud provider as part of a preferred vendor deal won't risk outright failure; they just risk losing money over time. For example, you might choose a preferred cloud provider for storage. But it might not be the best choice for database services, and it might only be partially optimal for compute services. All told, your enterprise could lose millions each year on lost productivity.
Enterprises who engage in this practice are quick to note that a particular cloud service might not be ideal, but it works. A business rarely tracks the cost of these "less optimal" cloud services, so it won't know what it's missing -- because it's not paying attention.