Walmart Plays Catch-Up with Amazon Using Machine Learning

Walmart uses machine learning to remain competitive in the retail industry, through personalization and process improvement at check-out.

 

Well known examples of retailers using machine learning include those in the subscription business, where the entire value proposition typically relies on a machine learning algorithm to consider users’ preferences in customization. However, more traditional retailers are also starting to embrace machine learning in different capacities, including but not limited to product development, inventory management, sales forecasting and supply chain management.

Walmart is a prime example, where employing machine learning for process improvement is critical to remaining competitive in the fast-changing retail environment. As the second largest online retailer, Walmart is competing head-to-head with online retail giant Amazon. Amazon has been using machine learning for years to provide and refine product recommendations on its website and in Kindle[1]. It uses purchase and browse history, as well as customer demographics and socioeconomic characteristics to provide relevant, customized product recommendations on an individual basis. Walmart recognizes that any failure to meet customers’ changing expectations while adjusting for technology investments and developments in its competitors operations could place them at a competitive disadvantage[2].

As such, if the company wants to maintain its position in the eCommerce space, while maintaining a significant brick-and-mortar presence, the most important near-term applications of machine learning will include optimizing the customer experience online and in mobile to increase conversion through personalization, as well as employing the technology to improve processes that directly affect the customer. Laurent Desegur, Vice President of Customer Experience, has echoed this priority, stating that Walmart is using machine learning used to create a model of personalization and recommendation1.

In the short term, in addition to using customer data to build a model of personalization and recommendation for its customers through optimizing product recommendations on site and mobile. It is also using it to remove bottlenecks at check-out, and is the technology behind self-serve kiosk towers, drive-up delivery and mobile check-out1.

In the longer-term, Walmart plans to develop a “cloud factory” to accelerate digital innovation. Specifically, it will leverage its recently announced partnership with Microsoft, aimed at enabling the companies to collaborate on machine learning and other forms of artificial intelligence[3]. Many applications of the technology will come from co-innovation around big-data and real-time analytics. For example, Walmart will be deploying Microsoft machine learning to launch a language processing platform to make sense of up to 40 terabytes of unstructured text. This will provide insights and educate business decisions in real-time[4]. Additionally, operating on a global scale on Microsoft’s cloud-based platform, Azure, will enable machine learning when routing thousands of trucks in the supply chain[5].

Because the essence of machine learning involves identifying insights from big-data, the technology is only as good as the data a company has, and as the people who interpret it. Therefore, in order to optimize the use of this technology, Walmart should emphasize quality as well as quantity in the collection of data. The company could provide incentive for customers to create and login to an account, where they will store personal information, preferences and other relevant data points. As such, Walmart can more accurately collect data – including click, browse and purchase data – on individual customers, all while providing a better customer experience through personalization.

In addition, while we can rely on automation to some degree, human intervention and judgement remains critical. Human judgement comes with its own set of biases that machine learning can overcome to a certain degree, however, the human element is important to ensure that critical thinking, and cross-validation is used when interpreting the data[6]. Walmart should invest in hiring highly-skilled data scientists to properly and effectively use the predictive technology.

With so many potential uses of machine learning in the retail space, how should Walmart prioritize applications of the technology going forward? Should it continue to focus on customer-facing process improvement, or shift to internal processes like forecasting and supply chain management? Further, how can it employ machine learning to gain competitive advantage, instead of using it to play catch-up with competitors like Amazon? (767 words)

[1] Paredes, Divina. “Machine Learning Is the ‘New Normal.’” CIO (13284045), Apr. 2018, p. 8. EBSCOhost, ezp-prod1.hul.harvard.edu/login?url=http://search.ebscohost.com/login.aspx?direct=true&db=bth&AN=129130726&site=ehost-live&scope=site.

[2] Walmart 10K: http://d18rn0p25nwr6d.cloudfront.net/CIK-0000104169/a25e7acb-aa07-49f3-8c0c-0c69e5a8d372.pdf

[3] Yeomans, Mike. “What Every Manager Should Know About Machine Learning.” Harvard Business Review Digital Articles, July 2015, pp. 2–6. EBSCOhost, ezp-prod1.hul.harvard.edu/login?url=http://search.ebscohost.com/login.aspx?direct=true&db=bth&AN=118667151&site=ehost-live&scope=site.

[4] Ho, Vanessa. “A New Walmart ‘Cloud Factory’ Will Accelerate Digital Innovation, Boost Business Efficiency.” Stories, 2 Nov. 2018, news.microsoft.com/transform/new-walmart-cloud-factory-innovation-business-efficiency/.

[5] Walmart Media Relations. “Walmart established strategic partnership with Microsoft to further accelerate digital innovation in retail.” July 16, 2018.  https://news.microsoft.com/2018/07/16/walmart-establishes-strategic-partnership-with-microsoft-to-further-accelerate-digital-innovation-in-retail/

[6] Yeomans, Mike. “What Every Manager Should Know About Machine Learning.” Harvard Business Review Digital Articles, July 2015, pp. 2–6. EBSCOhost, ezp-prod1.hul.harvard.edu/login?url=http://search.ebscohost.com/login.aspx?direct=true&db=bth&AN=118667151&site=ehost-live&scope=site.

 

Previous:

“Brewing a Better World” with Open Innovation: Heineken’s Innovators Brewhouse

Next:

An Epic use of Machine Learning

1 thought on “Walmart Plays Catch-Up with Amazon Using Machine Learning

  1. I completely agree that Walmart must use new technologies to remain competitive in the retail market. Machine learning is seen as a key to both increase customer satisfaction and optimize internal processes [1].

    I believe that Walmart should focus in both customer-facing process and internal processes. To have a competitive advantage it is important to increase customer experience, but at the same to reduce costs and increase efficiency of internal processes. Although it is true that the main focus of Walmart has been the customer experience, it has also developed initiatives that affect its internal processes. To illustrate, Walmart has introduced robots to facilitate the process of restocking the shelves in the supermarkets. It has been proved that the pilot not only improved customer satisfaction, but also enhanced internal processes by reducing inventory turnover and improving inventory accuracy [2].

    Walmart by operating more than 5,000 physical stores has to opportunity to use machine learning in the stores and have direct contact with clients. The self-serve kiosk tower mentioned in the article or the robot to replace inventories are some elements that Amazon cannot deliver to the customers, and then a competitive advantage of Walmart.

    [1] Dan Wellers, Timo Elliott, Markus Noga, “8 ways machine learning is improving companies´ work processes.” Harvard Business Review Digital Articles (May 31, 2017).
    [2] Retail Dive, “7 Ways Walmart is innovating with technology”, https://www.retaildive.com/news/7-ways-walmart-is-innovating-with-technology/525154/, accessed on November 2018.

Leave a comment