It’s no joke, "Person Drives 100 Miles in Wrong Direction, Following GPS" and “Fallout From The AP Hack's 'Priciest Tweet In The World” are both real news headlines, and strong reminders that automated processes still require the human touch.
The AP's Twitter (News - Alert) hack, and subsequent stock market dip, is just the latest evidence showcasing how and why next-gen technologies fail when humans are removed from the equation. While many speed traders followed automated processes, and reacted quickly to the false report of an attack on the White House, other firms held back, considering the tweet may not be legitimate.
In one case, a brokerage's managing director took to the company-wide loudspeaker in response to the news, saying, "Careful, those things can be hacked," warning his team to think before acting. Once the tweet was discredited, many traders emphasized the value of their human advantages – noting the person-to-person warning kept their organization from making costly and brand-damaging decisions.
Despite how automated the world becomes, no set of hard and fast rules – developed so far, anyway – can replace a human sensibility and intuition. Responsible companies that depend on accurate decision-making don't blindly trust algorithms to decide the fate of their customers' fortunes.
And that principle shouldn't just go for financial institutions. The same rules should apply for companies that value the fate of their customers' service interactions.
Machines May Be Smart, But Common Sense is Essential
Machine-to-human interactions are becoming more and more commonplace in customer service. Whether it's an interactive voice response tree or a much more comprehensive machine-to-human conversation fueled by natural language processing, companies and consumers alike now rely heavily on interactive technologies to bring resolution to needs via the channel of their choice.
But there isn't much room for error, and consumers have no patience for machines that are supposed to understand customer inquiries, but simply don't.
Thankfully, technology is rising to meet expectations. Customers are interacting with more intelligent virtual assistants every day – technology that is designed to engage in conversation, understand intent (based on customer input) and take correct action on behalf of users. By definition, these tools are built to understand context, and have the most advanced human-like intelligence behind them available.
Avoiding incidents like the AP hack and stock market dip are top of mind for companies implementing IVA technologies.
Infusing Humans into Automated Processes
In today's fast paced, high-touch world, scalable, 24/7 interactions through automated solutions are essential to providing top notch customer service. The human differentiators within these solutions – developed by both the engineering of the technology and in the process of escalating to live help – is where IVAs bring automated service to life, by functioning as trusted sources of information.
For a company like Next IT, which builds IVAs for organizations like the U.S. Army, United Airlines and Aetna, the human element has always been a part of the success of its overall IVA offering. By adding human inference to process and delivery, the IVAs are able to be taught over time to have the same sort of understanding and experience that makes interacting with people so easy and natural. But there are best practices to follow when checking for this critical human element.
1. Does the virtual assistant know the company's customers?
The process of being able to deliver the most intuitive experience for customers starts at the beginning of an IVA implementation process and is continuously monitored over time. Project experts should work alongside a company to discover and analyze common customer needs via existing data and subject matter experts.
2. Does the personality reflect the brand?
Some may remember Daimler's failed Dr. Z campaign in 2006, which drew heavy criticism for featuring a character to which consumers couldn't relate, or understand. Personality is a huge part of a virtual assistant's success – and a major opportunity to encapsulate, and bring life, to the brand. Personality should complement the functionality of the IVA and should include responses to the most commonly asked courtesy-related questions to maximize the stickiness of the assistant.
3. Does the virtual assistant know when to escalate to live help?
Based on business-established rules, some inquiries are decidedly best suited to be handed off to a live representative. Virtual assistants have to know when to pass the torch. Triggers for escalation must be appropriately timed, and be seamless.
Customer service and experience technologies need humans in the loop – through engineering, escalation and ongoing improvements – knowing when to ask for help. When corporations implement technologies that leave human intelligence out of the process, it's easy for a wrong decision or inaccurate answer to slip through the cracks. And while the price for those may not always be $136.5 billion, that underlying risk will always be there. Best practices and past experience tells us that the best way to avoid costly mistakes from happening is to keep sight of the crucial human element.
Mitch Lawrence is vice president of sales and marketing at Next IT (www.nextit.com).
Edited by Stefania Viscusi