Walmart's simply yuge AI spend................................................................................
" 55% of Amazon’s sales come from personal recommendations made by machine learning algorithms"
"A 2016 study from Retail Systems Research found that 59% of omni-channel customers were more profitable."
Everything goes better with AI chips.
Just the start.
Wal-Mart's AI Investment: Assessing The Possibilities For Revenue Growth
Oct. 2, 2017 10:35 AM ET
| About: Wal-Mart Stores, Inc. (WMT), Includes: AMZN, BABA, FB, FLWS, GOOG, GOOGL, KR, M, MSFT, NVDA, SPLS, TGT, WFM
by: LD Investments
Value, macro, long only
Get email alerts
Wal-Mart is reportedly planning to build a GPU farm about 10% of the size of Amazon’s. The investment is indicative of an overall investment in AI.
Strengthening its own AI capabilities could boost online revenue and support its omni-channel retail strategy.
Wal-Mart is also reportedly licensing out its cloud services to other companies, positioning itself as an alternative to AWS.
Please take note this is only one aspect in weighing the attractiveness or non-attractiveness of the companies mentioned as an investment and should not be used independent of other factors. This article examines one segment of the companies' businesses, and other factors such as valuation are not addressed. This article is not a recommendation to buy or sell any stock mentioned.
Wal-Mart’s (NYSE: WMT) AI investment could boost e-commerce revenue, support its omni-channel retail strategy and potentially emerge as a lucrative revenue stream which should ultimately help it better compete against Amazon (NASDAQ: AMZN).
Wal-Mart is reportedly building its own cloud-based data centers (expected to be about a tenth of the size of Amazon’s) using clusters of graphic processing units (GPUs) from Nvidia (NASDAQ: NVDA) to power custom deep neural networks which will be built on top of the GPU infrastructure according to Trip Chowdhry, a Global Equities Research analyst.
Nvidia’s GPU chips are being increasingly used for machine learning applications. Machine learning, which is just one of several technologies and processes that make up AI, refers to the science of self-learning algorithms where statistics and data from a knowledge discovery process is used to solve problems.
Wal-Mart’s investment in Nvidia chips is indicative of an aggressive push into AI. This effort could open doors to boost top-line growth.
AI Could Boost Wal-Mart’s Online SalesFor years, Wal-Mart’s e-commerce operations struggled while Amazon cruised upwards, taking an increasing share of the overall retail pie in the process. Last year however, following its acquisition of Jet.com and putting its founder Marc Lore in charge of US e-commerce operations, Wal-Mart has seen strong growth in online sales.
Source: Business Insider
As Wal-Mart ramps up its digital presence, an aggressive investment in AI could help Wal-Mart gain ground in the growing e-commerce market and catch up with Amazon. America’s largest online retailer Amazon, has invested heavily into AI and its AI capabilities give the company the advantage of being able to process copious quantities of customer data to provide online shoppers with targeted recommendations, thereby boosting sales by selling items that customers may not realize they need.
55% of Amazon’s sales come from personal recommendations made by machine learning algorithms while Target (NYSE: TGT) achieved 15% to 30% revenue growth with the help of machine learning algorithms.
AI could become a major differentiating factor in the future of retailing and AI technologies are increasingly gaining importance in the retail industry. A significant proportion of retailers are planning to invest in IoT and AI technologies in an effort to differentiate themselves in an increasingly competitive environment.
Source: Business Insider
Kroger (NYSE: KR), Target, 1-800 Flowers (NASDAQ: FLWS), Macy’s (NYSE: M) and Staples (NASDAQ: SPLS) for instance have invested or are planning to invest in AI technologies.
AI Could Drive Wal-Mart’s Omni-Channel Retail EffortThe retailing industry is moving towards omni-channel retailing, a concept which fuses e-commerce with physical stores, to create a seamless shopping experience for the customer. A 2016 study from Retail Systems Research found that 59% of omni-channel customers were more profitable.
Machine learning is expected to play a role in omni-channel retailing, particularly for the largest retailers that have the scale and financial muscle to implement the technology.
Very broadly speaking, there are two key requirements for machine learning: data, and machine learning infrastructure (which usually includes among other things, GPUs which are designed to process large amounts of data quickly and more cost efficiently). The more data is fed and the more times the data is analyzed by the system, the more accurate the resulting model is likely to be.
Amazon’s omni-channel ambitions may have been the driving force behind its acquisition of Whole Foods (NASDAQ: WFM) as doing do would provide Amazon with vast quantities of data on shopper behaviors at retail stores which could be used to study and train its own AI systems. 90% of worldwide retail spending still happens at physical retail stores according to eMarketer
Wal-Mart, with its unrivaled physical store network and its position as America’s second largest online retailer is in prime position to benefit from the omni-channel retail trend. Wal-Mart sees 100 million shoppers per week across its network of five thousand stores. Data from shoppers at its physical stores coupled with data from shoppers on Wal-Mart’s websites generates a massive quantity of data. These data silos have to be combined and analyzed in order to implement an omni-channel retail experience. With the help of machine learning, Wal-Mart could better utilize this vast quantity of data to train their AI systems which, if well implemented, could lead to improved sales.
A new source of revenue?However, there is reportedly more to Wal-Mart’s AI cloud ambitions than just improving internal business intelligence and boosting retail performance.
Wal-Mart has been a long time user of AWS (Amazon Web Services). However with Amazon beginning to compete directly with traditional brick-and-mortar retailers after its Whole Foods purchase, Wal-Mart (as well as other retailers such as Target) is reluctant to share their data with Amazon. As a result, Wal-Mart has been trying to move its data away from AWS by developing its own AI infrastructure rather than add to the revenue of its biggest competitor. Target is also reportedly planning to scale back its use of AWS.
Amazon derives the vast majority of its profits from its cloud computing division (despite the fact that AWS contributes less revenue than its retail division), and thus Amazon is effectively subsidizing the cost of its e-commerce business.
Wal-Mart appears to be taking a page from this business model and is reportedly licensing its cloud computing services to other companies, effectively positioning itself as an alternative to AWS. Wal-Mart has previously asked its suppliers to stop using AWS. With its new cloud business, Wal-Mart may be aiming to get those vendors on its own cloud instead.
Wal-Mart’s foray into cloud computing might slow down Amazon’s cash cow, however, it is unlikely to be a major dent for AWS which is currently the largest cloud infrastructure services provider with a market share of more than 34%.
Source: Business Insider
Nevertheless, it could offer Wal-Mart another way to compete against Amazon as well as an opportunity to diversify its revenue stream.
Amazon, like Google (NASDAQ: GOOGL) (NASDAQ: GOOG), Microsoft (NASDAQ: MSFT) and Alibaba (NYSE: BABA) built their massive data centers and software, not with the intention of renting it out to others as cloud computing providers but to further their own business goals. For Google it was about being able to process the three billion searches it handles daily. For Amazon and Alibaba it was about processing the thousands of orders per second on major sales days such as Prime Day for Amazon or Singles Day for Alibaba. In order to accomplish these feats, these companies were compelled to develop custom-built, tailor-made hardware and software to handle their unique workloads.
AWS was born as a way for Amazon to make extra money on the massive infrastructure the company had built to support its retail operations. From just one customer at the beginning (Amazon), AWS amassed an army of customers, emerging as a lucrative business segment for Amazon. Amazon is still AWS’s biggest customer.
Like Amazon and other tech giants, Wal-Mart maintains a substantial IT infrastructure to support its massive retail operations as the world’s largest retailer and America’s second largest online retailer. In 2013, Wal-Mart acquired OneOps, a cloud startup, and developed its own OneOps cloud to run much of its infrastructure. Both Wal-Mart.com and samsclub.com are managed with it. The OneOps purchase helped transform Wal-Mart’s e-commerce software from a cumbersome, monolithic application into a more agile one and Wal-Mart’s OneOps team is now spearheading the company’s AI initiatives.
Cloud computing is a fast-growing market and as demand for AI grows in the near future, Wal-Mart’s effort to develop its cloud offering could open the opportunity of capturing some of that demand.
While Wal-Mart is usually not mentioned in the same breath as tech companies such as Google, Facebook (NASDAQ: FB) and Amazon, Global Equities Research analyst, Trip Chowdhry advises that “investors should not underestimate the Technology acumen of Wal-Mart Software Developers,” adding, "we have seen them present at various conferences, they are as smart as Google or Facebook engineers."
Interestingly, Wal-Mart, which has a reputation for paying barely above the minimum wage to its thousands of low-skill floor employees, is actually paying a lot of money to its software engineers, thereby placing it within the list of 25 highest paying companies in America. A recent Glassdoor survey found that software development engineers at Wal-Mart can average as high as US$ 151,666 a year, higher than the average at Amazon where software engineers are paid an average US$ 131,885.
How popular Wal-Mart’s cloud service remains to be seen, however, if successful, it could potentially be a lucrative revenue stream for the company.
To conclude, Wal-Mart’s investment in Nvidia’s GPU chips to build a GPU farm expected to be about 10% of that of Amazon’s is an indication of an aggressive push into AI. AI technologies such as machine learning are expected to be key differentiators in retail going forward and several large retailers that have the scale and financial muscle to afford the technology have invested or are investing in AI. For Wal-Mart, such an investment could help boost revenue in its e-commerce division for instance through better product recommendations which should allow it to better compete with Amazon which has harnessed the power of machine learning to generate over half of its revenue. As the retail world moves towards omni-channel retail, Wal-Mart with its unrivaled store network and e-commerce business which is the second biggest in America is in prime position to benefit from the fusion of online and offline shopping. Beefing up its AI capabilities better allows the company to process and utilize the data as part of its omni-channel strategy. Finally, Wal-Mart’s AI initiative appears to be aimed not just for internal business intelligence and improved retail performance; Wal-Mart is reportedly licensing its cloud computing infrastructure to other businesses, positioning itself as an alternative to AWS which is Amazon’s most profitable business. Although Wal-Mart’s entrance into the cloud game is unlikely to cause significant damage to market leader AWS, if successfully implemented, it has the potential to emerge as a lucrative revenue stream for Wal-Mart.