Forbes calls the Singles Day shopping festival “like the Academy Awards, a New Year’s Celebration, and the Super Bowl all in one.” Indeed, the Alibaba-invented “retail-tainment” on November 11 drew in consumers to shop to the tune of $25 billion. Compare this to the paltry $3 billion spent during the US-centric Black Friday and Cyber Monday. The Chinese mobile payment system Alipay was in full force, accounting for the vast majority of the Singles Day sales, compared to only about 30% of shoppers in Black Friday and Cyber Monday last year.
While large-scale site meltdowns—think “page not found”—are mostly a thing of the past, due to major precautions retailers take, such as massive load tests, the threat this year has been the multiple, harder-to-identify microglitches. As online commerce and internet and mobile apps has grown exponentially, so has its complexity. Glitches have scaled down in size but grown in number and sophistication. Mispriced items, missed revenue opportunities, and broken customer journeys all have a cumulative effect on revenue and brand, often more insidious than catastrophic failures.
When people think of artificial intelligence applications for ecommerce and retail, they most often think of things like personalization and recommendation engines that interpret a consumer’s past purchases and behavior and then tailor their next shopping experience appropriately. Yet these types of applications have a completely different time scale than is required to identify microglitches. Data updates to a personalization algorithm can be made on a daily or even weekly basis. To identify microglitches, real-time continuous learning AI is needed.
Shyam Sundar explains how to use unsupervised machine learning to keep websites and mobile apps running smoothly under the stress of massive numbers such as those seen on Singles Day. Along the way, Shyam shares examples of microglitches and business opportunities facing online and mobile businesses and discusses how AI can be a key part of an online strategy for identifying them in real time. By collecting time series data metrics from multiple sources, such as web traffic, servers, weather, third-party ad bids, social media metrics, in-store transactions and so on, and learning their normal behavior with online machine learning, it is possible identify when any metric behaves abnormally. AI can further be used to correlate among multiple metrics, triggering an alert when, for example, revenue drops for a particular product at the same time a competitor significantly increased it advertising bid for the same product. With this method, pricing errors, conversion problems, and business opportunities can be caught early and resolved, protecting companies against revenue loss and brand damage.
Shyam Sundar is the Sydney-based regional director of APJ at Anodot, a leading provider of AI-powered analytics. Shyam has 17 years of experience in strategic data innovation, advanced business intelligence, analytics, and anomaly detection. Previously, he worked in APJ for companies including Cloudera, HP Vertica, Sybase, and numerous startups. Shyam has also consulted on cutting-edge data strategy for Global 2000 companies. Shyam holds a bachelor of engineering from Annamalai University in India and an MBA from Monash University in Australia.
©2018, O'Reilly Media, Inc. • (800) 889-8969 or (707) 827-7019 • Monday-Friday 7:30am-5pm PT • All trademarks and registered trademarks appearing on oreilly.com are the property of their respective owners. • email@example.com