Walmart’s Edgy Strategy
By: Noah Rawitz
Barely a month has passed since I took my Business Strategy final, and the Wall Street Journal is already providing me a fantastic opportunity to apply classroom frameworks for analysis— I feel validated. Check out this diagram from an article about Walmart’s “secret” strategy plans to compete with Amazon and see if you can spot the components of #AAA :
For the uninitiated, according to the AAA framework, a firm can create value by utilizing its Assets to perform Activities which lead to a competitive Advantage (and if you are jut now becoming acquainted, you owe me tuition). In the diagram above, Walmart’s supercenters are its key assets, and each of the spokes represent activities that leverage these assets to create value. Like all forward-thinking MBA students, I focused on the upper-right “Planned” quadrant of the diagram, and one activity caught my eye — Edge Computing. From the article:
Turning far outside its retail expertise, Walmart plans to build “edge computing” capacity, in which data is processed physically close to where it is being collected, a faster system than sending data to the cloud. That system, spread out among retail locations, could be rented out amid new demand by autonomous vehicles and other systems that may use the technology to process large amounts of data quickly, according to people familiar with the company’s plans.
Although the article makes it clear there is internal skepticism of the success of such a tactic, the fact that Walmart is even considering edge computing got me thinking I should, too.
What is edge computing?
Most are familiar with the basics of how cloud computing works: 1) Data is collected locally 2) Data is sent off to centralized location to be analyzed/processed 3) Data (and often decision-making “commands”) is sent back. The “cloud” is quite literally a collection of servers in a warehouse (in Amazon’s case, these servers are housed in Virginia, San Francisco, Seattle, and Oregon). Edge computing, in contrast, moves the processing in step 2 physically much closer to the actual devices that need to take action using the returned data. Here are a few main benefits of doing this:
- Reduced Latency: The average cubicle worker can afford to wait a few seconds for data to be returned (although I’d still recommend sighing loudly at your desk while a dashboard loads so your office mates know how hard you’re working). But for self-driving vehicles, real-time interpretation of sensor data could be the difference between life or death. Even for manufacturing companies, real-time analysis can increase operational efficiency (imagine a machine part that intelligently forecasts deterioration before failure, saving tens — or hundreds — of thousands in downtime). We’re talking about the capability to process data in microseconds as opposed to milliseconds.
- Lower Costs: Cloud computing and internet bandwidth can be expensive, especially as data ingestion and collection grows exponentially for nearly every company. Edge computing allows companies to be more selective when deciding what and how much data gets sent to the public cloud.
- Improved Security*: Because data is stored locally, edge computing may be less vulnerable to cyber attacks, as there is no single point of access for hackers. *Security is a benefit in theory but may not be in practice. More on this later.
Is edge computing a good strategy for Walmart?
It depends! (Again, for the uninitiated, “it depends” is always the correct answer in a business strategy class).
There is no question that edge computing is going to be important in the future (particularly for IoT and self-driving technologies), and the market has some big dollar signs attached to it. Grand View Research estimates the global edge computing market will be worth over $28B by 2025. Major cloud providers are getting ready. Partnerships are forming with telecommunications companies to leverage 5G to expand their mobile edge capabilities (Amazon+Verizon / Microsoft+At&T), and many are also offering other edge computing services. So, it’s smart of Walmart to take notice. It’s even smarter for Walmart to realize that its network of nearly 4,800 retail locations in the United States (asset!) provide it a natural head start in local computing resources (particularly compared to Microsoft and Google). Self-driving companies hungry for edge processing could feasibly turn to Walmart’s supercenters to power their technology across the country.
That said, the question is whether or not Walmart can execute on this strategy successfully. First, this means actually developing the technology, which is no small feat. Walmart would have to hire and train cutting-edge (ha!) technologists and would have to compete with the bigger, sexier players when doing so. Walmart would also have to commit to maintaining and upgrading their technology to keep pace with innovation in this rapidly growing space. Second, Walmart would have to do this in a cost-effective way if they hope to capture any of the revenue as profit. In addition to the associated hiring and development costs just mentioned, maintaining a secure data center of any kind is also a significant investment (Amazon more or less equates their centers to Fort Knox). Security can’t be overlooked for the full value prop of edge computing to be realized.
Perhaps Walmart’s best option — as this Forbes article offers — is to partner with a major cloud provider. In that world, the two companies would be able to capitalize on both real estate and computing expertise: win-win. #synergy.
