By Paul Mandeville, QuickPivot
In its recent forecast about cognitive technology, International Data Corporation (IDC) says spending will reach $46 billion by 2020, a seven-fold increase over 2016’s levels. Artificial intelligence is pushing into everyday business reality, and customer experience management is a chief driver of the growth. Chatbots, it logically follows, are the way of the future.
Not so fast. Artificial intelligence should not be taken as a be-all, end-all answer to customer service. To be sure, cognitive technologies introduce a subtle, yet perceptible, lack of empathy; they pose increased risk to brand credibility; and they could become an accelerant for IT security woes.
Advertisement
Lack Of Empathy And Authenticity
Customer experience is highly emotive; people who engage with a brand through messaging want to believe they’re having an actual conversation with an actual person. The hard truth is, chatbots simply aren’t sufficient to uphold this desired value. They exhibit passive response cycles, are limited in their dialogue capabilities and, worse yet, are capable of brand-eroding meltdowns. At the end of the day, chatbots are the marketing equivalent of the “uncanny valley,” where animatronic human replicas that are nearly, but not quite, like real human beings elicit the strongest feelings of revulsion. You do not want to unwittingly induce a dip in affinity for your brand!
Customer interactions cannot, and should not, become so automated that businesses lose sight of the purpose behind human touch in the first place — empathy and relationship building. Machine learning has many high-value business use cases, but customer engagement just may not be the right venue. Indeed, “automation” and “customer service” may be oxymorons.
Loss Of Credibility And An “Experience” Disrupted
AI chatbots are largely reliant on a carousel of pre-scripted responses, which are in turn based on a given set of keywords. But if the bot does not recognize the keyword it’s being thrown, the interaction is at risk of meeting a melodramatic demise, and bringing the customer relationship down with it.
Courtesy of The Ringer
Nonetheless, a 2016 study by ubisend showed more than half of consumers believe businesses should be able to respond to their queries 24 hours per day, 7 days per week. There is a definite tension between the need for ubiquitous response and the challenge of providing human talent to meet that demand.
Most people just want to know they are being attended to, so use the chatbot’s “talents” in the way they are meant to be used: For quick, single-answer, fundamental questions. Be prepared, however, for a next line of inquiry that requires a more complicated, personal response. The mere act of communicating a clear course of action is a crucial — and often neglected — step toward building a credible, one-on-one experience.
Accelerating Insecurity
Ironic as it may seem, the criminal underground has leveraged chatbot tech for years, where the age-old “botnet” has been a go-to deployment vehicle for nefarious cyberattacks. “Denial-of-Service” attacks are effectively a chatbot at work.
AI systems are uniquely susceptible to bad human actors armed with misleading data that could teach them to make decisions, or take actions, that hurt the brand. Owing to the chatbot’s nature, this activity can go on undetected for long periods of time. We need only look to high-profile examples such as the Microsoft Tay fiasco, Tinder and the DNC to understand just how deep the ramifications of undetected actors can reach.
When a breach occurs by way of AI, it can reach conflagration status quickly and be much more difficult to trace and resolve.
AI: The “Person” Behind The Person
In a word, cognitive computing is best suited to become the human’s perfect partner in providing customer experience. As a recent HBR.com column suggests, AI collaboration resides on a spectrum, ranging from “assistant” to “colleague.” As a colleague, the chatbot brings to the table a continuously data-driven grasp of situations that the human brain has difficulty sustaining. Bots can spot issues more quickly, help humans prioritize their activity in the most rational way, and serve as a sounding board to quickly decipher a customer’s end objective.
While smartly divided labor can create better knowledge, inspiration and creative exploration across the entire marketing workforce, in a world that strives to create a great experience for each and every customer, humans matter. And they always will.
Paul Mandeville serves as Chief Product Officer of QuickPivot, leading the firm’s efforts in product strategy, innovation, and design. Prior to joining QuickPivot, Paul served as Chief Operating Officer at Conversen, a cross-channel marketing technology start-up in Boston. He has more than 15 years of marketing technology experience, holds several MarTech industry awards, and contributed to QuickPivot’s first U.S. Patent grant for the design of QuickPalette.
