E-Pulse – AI and Innovation in E-commerce
Think Tank: With Personalization, Retailers Will Fix Big Data Misses
For retailers, big data represents the marketing renaissance that never arrived. Data science made remarkable strides over the past decade, with retailers eagerly investing millions in big data tools as a means to multiple ends, from improving omnichannel shopping experiences to integrating logistics merchandise, and payment tools. And indeed, some gains were achieved operations, supply chain and logistics. Yet, many department stores and malls are struggling. Chains like Sears Holdings Corp., Macy’s Inc. and J.C. Penney Co. Inc. are shuttering 100 or more stores alone. The Limited closed all 250 stores and cut thousands of jobs. Gap Inc. has also seen sales declines in 10 of the last 11 quarters. Abercrombie & Fitch closed 60 stores last year, and 200 in the last two years.
From a market competitiveness standpoint, big data was intended to forestall these closures, providing retailers with meaningful, actionable ways to understand customers’ personal styles and interests. Instead, it has given rise to a cheap form of “personalization” based primarily on redirected ads and past clicks, not a true understanding of the consumer.
The potential payoff from true “personalization” has yet to be realized. A recent survey by Time Trade Systems, a customer experience analytics company, found that consumers are willing to increase their spending by about 4.7 percent in exchange for better in-store experiences, including more personalized service. That extra spending adds up to $150 billion in unrealized revenue, the researchers found.
But improving the customer experience takes more than big data, it takes personalization technology and predictive taste recognition tools powered by next-generation technology.
“The potential payoff from true “personalization” has yet to be realized. […] That extra spending adds up to $150 billion in unrealized revenue”
Fortunately, retailers can now use artificial intelligence, and specifically deep learning tools that construct neural network maps, bridging the online and in-store activity of individual shoppers and generating unprecedented levels of personalized customer support. This is made possible through the use of a wider range of inputs — including images of products that shoppers browse inside stores and scan for more information — than legacy personalization tools.
And over time, as a deep learning platform accrues more data on individuals, its accuracy, in terms of prediction, increases.
Past Is Not Prologue
We’ve all walked into a department store and interacted with a salesperson who, by some unexplainable alchemy, seems to understand our tastes with uncanny speed and deftness. He or she suggests the perfect apparel pairings. It seems like the salesperson is a mind-reader.
But top department store sales associates are not clairvoyant, they’re well-trained, and they’ve honed their skills of perception over time. They are also a major, and scarce, asset for department stores.
Now, imagine you’re in that same department store with a salesperson suggesting apparel. The sales associate brings you an item you just tried on, something you already purchased, or worse, something that’s a complete mismatch with what you’re looking for. It’s likely you won’t return because it’s obvious that the salesperson doesn’t understand you or your style. Online, we’re experiencing a similar ineffective and frustrating approach.
That’s because the majority of e-tailers rely on a “If you bought this, you’ll also buy that” approach that is too focused on the past as a roadmap to future purchases. This is the conditional algorithmic approach mastered by Amazon and the 300 million sku’s and data points it collects, but it’s clear that this approach cannot account for nuances in taste. Our personal tastes evolve over time — they’re not just influenced by quantifiable data from marketing campaigns and ads.
A New Personalization
For retailers to succeed, leveraging big data toward a new kind of personalization — one that predicts your next purchase based on a broader range of inputs and nuances — is paramount. Shoppers are unpredictable, and the window of time during which retailers can turn interest in a given product into a transaction is also highly variable. But just as e-commerce has given consumers tremendous access to product, via comparison — and omnichannel-shopping, digitally-connected sales associates can better connect shoppers with products they’ll love.
Until now, it’s been possible to separate expectations between online and brick-and-mortar levels of personalization: traditionally, in-store would always be better. But shoppers want a seamless shopping experience: what they find online should match what they find in-store. Some retailers are already taking steps to bridge the gap.
Neiman Marcus is adding a human touch to its mountains of consumer data. The luxury retailer is looking to its sales associates, those closest to shoppers, to gather human data: observations about how customers shop, what they say and their preferences. Those insights will be translated to data, delivering a more human marketing experience for its shoppers. Others are also doing similar things.
Hudson’s Bay Company, for example, with its $250 million acquisition of Gilt Groupe, is focusing efforts on data-driven e-commerce and personalization. Another example is Nordstom with its acquisition of Trunk Club, an e-commerce site that is allowing users to sample looks at home. The North Face is also using artificial intelligence to drive e-commerce sales, using IBM’s Watson to engage with customers and serve personalized recommendations based on user preferences.
New tools for personalization are also emerging. Neural networks — like a computer system molded after the human brain — coupled with image-recognition technology have a supreme effect for retailers looking to capture the most actionable data possible because they analyze how people shop. These tools strive to understand taste by detecting thousands of elements of an image, using AI and machine-learning to adapt to customer preferences over time.
To succeed in e-commerce, retailers need to create more personalized experience. By offering a predictive nature to data analysis, retailers are provided with actionable insights that give value today — not in the coming months. This is how retailers, facing declines in brick-and-mortar stores, create community around brands, building loyalty by personalizing the shopping experience.
What’s your take on this? How do you see retailers delivering a personalized experience?
The original article was posted on WWD
Get in touch for more information.