Can we put our fate in a chatbot’s virtual hands?

September 20, 20175 Minute Read

Select article text below to share directly to Twitter!


Are chatbots for real? If you believed the tech industry hype, you might think chatbots were totally ready to help you live your best life, assisting you with everything from buying flowers for your mom to booking a rental car for a summer weekend getaway.

But is that rosy outlook about the use of chatbots realistic? Let’s take a measured look at chatbots, why they’re getting so much buzz, and whether they actually make sense for businesses right now.

Wanted: Chatbots ready for action

These days, consumers expect a personalized response to their questions—on the spot, 24/7, and with as little friction as possible. That means no droning customer service reps asking you to answer the same rote questions over and over before you actually get some help, no wandering in the wilderness of the soul-crushing corporate phone menu, and no waiting endlessly for a response to an email.

When faced with those archaic customer service channels, consumers would much rather conduct transactions using an efficient messaging app. In fact, according to ubisend’s 2017 chatbot statistics, 45.8 percent of consumers would rather communicate with a business through a messaging app than through email.

Given how new messaging apps are on the scene, that statistic might come as a surprise, but when you stop and think about it for a minute, it starts to make sense. We’re increasingly using short, quick messages to communicate with almost everyone in our lives—whether that’s via a messaging app, text messaging, or social media—and we want to coordinate our business transactions the same way, too. Given this sea change in the customer service landscape, it’s no wonder businesses are exploring chatbots as a way of scaling up their customer service operations to meet these increased expectations for constant availability. One day, chatbots might empower businesses to do just that—without breaking the bank.

Useful chatbots on the scene

As much as some of us may welcome our AI overlords or fear the wrath of Skynet, we’re not living in the future yet. So, what’s the business use of chatbots actually like today? Chatbots are in the perfect spot to take on customer service roles by relieving human representatives from the burden of fielding tedious and repetitive inquiries. In this respect, chatbots fall right in line with other forms of automation—assuming the burden of simple tasks so busy professionals can concentrate their time on higher value projects.

As VentureBeat notes, chatbots already help customers place orders for food online directly from Facebook Messenger. They’re good with follow-up, too: When ordering food on a rainy night, chatbots can check in via text message after the delivery to gather customer feedback and ensure that the experience was positive. In both cases—particularly the second one—chatbots are beefing up staff capacity and allowing businesses to operate on a lean staffing model, while ensuring a quality customer experience and remaining competitive.

That’s not all chatbots are up to, though. Businesses have also deployed them in e-commerce apps to help customers pick out the right outfit and answer frequently asked questions about store policies. They can accomplish this via multiple channels at once, which is something humans find difficult to do since our brains were not designed to multitask—no matter how much we may protest to the contrary.

Chatbots are even learning how to mimic human negotiation patterns. According to Recode, Facebook’s bots are trained to negotiate on behalf of businesses. Human reps usually take the wheel with this one, particularly when trying to prevent subscribers from canceling or attempting to steer customers toward a new product or service that could generate revenue for the company. Although these chatbots are still under development, it’s clear even now that they could one day become an integral part of how businesses engage with their customers.

When bots drop the ball

To say that chatbots are ready for prime time is a stretch. We’ve already had one eyebrow-raising case in which a seemingly harmless chatbot went off the rails—and fast. Microsoft ended up with egg on its face after its chipper AI-powered Tay chatbot learned some bad language from Twitter trolls, suddenly went rogue, and started issuing inflammatory tweets before being swiftly taken offline. Although Tay was more of a proof of concept than anything else, her sudden demise demonstrated that automation gone wrong can rapidly cause more problems than it fixes.

Although most chatbots don’t flame out quite so spectacularly, they do frequently show their limitations, because frankly, they’re not that intelligent yet. If a chatbot is only programmed to handle specific queries, customers with a bit of snark can quickly get it to issue a nonsensical or brain-dead reply. Then, it’s back to business as usual, with customers shouting, “I NEED A HUMAN” at the automated services. This obviously rules out sensitive or risky applications, like automated legal or medical advice, as MIT Technology Review notes, and it seems to narrow the range of tasks for which chatbots could be successfully dispatched—at least for the time being.

One day in the not-too-distant future, though, chatbots will have the smarts necessary to navigate interactions with customers—with far greater sophistication and efficiency. During that time, businesses will get a better handle on which types of business processes work best for chatbots. We don’t want to set them up to fail, obviously.

Artificial intelligence for IT tasks could be just around the bend. Once these virtual concierges are trained to navigate the complexities of human communication, businesses may give them the green light to enter the workforce en masse. Are you ready?

  • Recommended for you
  • Recommended for You