Whole Foods under Amazon will undergo a digital transformation, its data put to new and better use. Enterprises adopting hyperscale go through a similar change.

In 2008, I was in charge of the media planning department in a boutique agency in New York. I had been brought on board mainly to handle one particular client: Amazon. They were preparing to launch their first e-book reader, the details of which were so secret that we weren’t supposed to say its name aloud in the office. Whenever I see one, I still think of the code name: Fiona.

Apart from the K---I mean Fiona, Amazon was exploring a few other business channels, testing the waters for larger future initiatives. One of those channels was informally referred to as Amazon Grocery. There was no intuitive URL for Amazon Grocery, but if you knew where to click (or if you clicked on one of the ads that my group deployed), and you lived in one of the test markets, you would end up on a page at amazon.com that would offer to sell you groceries and other household items that Amazon wasn’t typically offering at the time.

Why is Amazon interested in groceries?

My colleagues and I often speculated on what Amazon Grocery was really about. Why was Amazon interested in food? Was the idea to compete with FreshDirect? Were they going to build special warehouses to store perishable goods? One day, the ads stopped running, and Amazon Grocery disappeared as if it had never existed. Since then, I’ve occasionally wondered what the point of it was. But now I have a pretty good idea.

A few days ago, Amazon announced that they were acquiring Whole Foods. At first blush, the two organizations seem to have almost nothing to do with each other (except for an overlapping base of early adopters). But a closer look reveals some interesting connections between this acquisition, Amazon Web Services (AWS), and cloud computing in general.

Amazon builds a service-oriented architecture

Whole Foods does one thing very well: it provides upscale healthy food in a supermarket environment. Amazon has gone from doing one thing very well (selling books) to managing one process very well: distribution. Along the way, they have created a number of autonomous platforms that can be inter-linked or disaggregated. Each of those platforms has, in turn, been transformed into an outward-facing product, honing its efficiency through competition on the open marketplace.

In a sense, the Amazon-Whole Foods contrast is the contrast between a more traditional organization, and a company that has gone through a digital transformation--that is, it uses its data to make money. No doubt, Whole Foods has a lot of data on its customers and products, and its whole value proposition is that it’s more progressive than the average grocery store. However, at the end of the day, it’s still just a chain of specialized stores.

In traditional datacenters, it's common to find groups of resources that have become highly specialized at doing one thing, and are too expensive to replace, but also not flexible enough to do anything else. Those components might go very well together, and they may have been built by manufacturers that had a particular type of workload in mind. But during a down period, that equipment is wasted.

The future solution is highly efficient software-defined hyperscale infrastructure (SDI), which allows for the creation of virtual performance-optimized datacenters (VPODs). VPODs are flexible, pulling the components that they need from resource pools, and allowing unused components to contribute wherever they’re needed. But what does this have to do with Whole Foods?

How Amazon and Whole Foods relate to cloud infrastructure

Well, now that Whole Foods has been acquired by Amazon, it is no longer an island but part of a collective. Its data becomes useful, not just in selling more health food at brick-and-mortar locations, but to help Amazon learn how to sell all sorts of other things to the same customers. And Whole Foods no longer needs to be profitable by itself to survive, which means that it can be a laboratory in which Amazon can learn how people shop when they’re not online. In other words, Whole Foods represents a collection of resources that can be used for multiple functions, and are therefore less likely to be wasted.

Amazon went through its own digital business transformation with the creation of Amazon Web Services (AWS). In creating AWS, Amazon took the flexible data distribution network they had built for their own purpose, and made it customer-facing, effectively turning a cost center into a profit center. And Amazon has continued to do that with other parts of itself over time, all along its vertically integrated supply chain, leveraging its autonomous platforms.

Amazon Grocery, as it turned out, was a lot like a vPOD. It lived for a short time, it made use of existing resources that had been previously used for other projects, it served its purpose (to collect data), and then it was dissolved. Perhaps, one day, it will live again. After all, the resources are available.

Learn how elastic vPODs - part of Ericsson's implementation of software-defined infrastructure based on Intel® Rack Scale Design - can make your architecture agile and lean.

Download the paper


Build your cloud and NFV infrastructure Inspiration & knowledge

Michael Bennett Cohn
Follow Michael Bennett Cohn:

Michael Bennett Cohn

Michael Bennett Cohn is a digital content expert with a wide range of experience in online publishing and advertising. He was head of digital product and revenue operations at Condé Nast, and he oversaw the digital ad campaign that launched the first Amazon Kindle. His journalism has been published by Newsweek, Mashable, and The Forward. He lives in Brooklyn.