Hitmetrix - User behavior analytics & recording

Robots Are Taking Over Retail

Though not quite as exciting (or terrifying) as artificial intelligence per se, chatbots and virtual assistants are proliferating, especially in retail. Through a partnership with IBM Watson, Macy’s recently joined the growing list of retailers that are implementing AI-infused technology to help improve customers’ shopping experiences.

That brands are flocking to Watson-esque technology is more sensible than it is remarkable. Retailers have long struggled to meaningfully marry digital with their in-store business, and while many assumed the advent of Big Data would help address some of the divide here, processing and actioning these reams of data has proved a considerable challenge for many. So, enter the robots; machines capable of quickly processing company and customer data and conveying their learnings to customers, often in a natural language conversation.

Obvious benefits aside, the increasing adaptation of Watson and its ilk begs a couple of questions: how are these chatbots and virtual assistants actually fitting into customers’ everyday shopping experience, and (crucially) what role will human customer service agents play in a retail world ran by all knowing, talking/texting machi1nes?

To address the first question, it’s important to understand one of the top reasons marketers are taking the assistant route in the first place. “Today, according to Forrester, 84% of consumers are self-serving on the company website. All we’re trying to do here is improve that experience,” says Marina Kalika, senior director of product marketing at TouchCommerce, an online engagement solutions provider. Kalika notes that more than 50% of customers abandon a purchase if they can’t find answers on the website, and 40% of them will end up making a call to the brand’s call center at that point.

In effect, the influx of bots, and AI-infused technology like Watson, is a direct response to consumer behavior. Whether people make use of the various brand bots en masse remains to be seen, but the success of virtual assistants like Siri and Google Now (soon to be Google Assistant) can be viewed as something of a proof of the concept of consumers regularly interacting with AI and contextual machine learning technology. But this brings us to our second, and perhaps more urgent dilemma: where does this leave human workers? More or less where they are now.

“Watson will not replace human customer service agents,” says IBM Watson Platform Manager Jonas Nwuke. “Cognitive systems like Watson augment human intelligence, working side-by-side with humans to accelerate and improve decision-making. It will give them an extra tool at their disposal to help them provide better quality service for more customers.”

Even more than that, retailers will struggle to maintain and improve their chatbots without human intervention. Marketers need to work with their technology partners to ensure their data is clean and accurate, and that the bots are referencing that data correctly. Humans will prove essential in moderating the chats brand chatbots are having as well, lest Microsoft’s Tay fiasco repeats. And if these tasks sound specific to engineers, consider what this conversation about bots means for  customer service.

The logical progression of customer service in a self-serve economy is likely in live chat, a service many retailers offer on their websites today. These systems function similarly to the messenger-based chatbots proliferating in the market today. Users open the chat, and are often greeted with an automated message (a bot) before continuing the discussion with a human representative. If the current trends in machine learning and natural language continue, it’s perfectly conceivable that these live chats will be ran by sophisticated virtual concierges.

Rather than absorbing the existing role of customer service agents, human representatives could find themselves the stewards over these chatting machines; no longer charged with manning a phone, but with facilitating the transfer of customer information, and moderating the bot’s discussions.

Related Posts