Last week I wrote about the first hot customer service trend for 2017, the Internet of Things (IoT). Today, I want to share some expert predictions on another hot commodity for customer service in 2017: Bots.
We know that small businesses can cost-effectively take on larger competitors by beating them at customer service. Chatbots and other virtual assistants can be a great way to supplement your customer service offerings in a cost-effective way.
“Bots can deal with all the repetitive questions sent to customer support teams, which frees up support staff to answer much higher value queries,” writes Paul Adams, VP of Product at Intercom. “The effect of this will be that support teams will shift from being perceived as cost centers, to being seen as increasingly strategic assets to successful companies.” Bots can help replace apps as a distribution channel as the end of apps and the end of app stores approaches.
In 2017, we’ll see more and more brands jumping on board. By 2019, use of virtual customer assistants will triple, according to Gartner.
GetApp covered the Gartner Symposium 2016, and one takeaway was that “Your next customer service assistant will be virtual.” While virtual assistants made an appearance at the 2015 Gartner Symposium, they were much more prominent this year. During the keynote, virtual assistant Amelia helped Gartner analysts on stage. And in a breakout session on the future of work, Gartner analyst Diane Morello predicted virtual doppelgangers will take our place in meetings, freeing us up to be more productive. Your lips to God’s ears Diane!
A robot to go to meetings for you might seem far off, but try to buy a phone without AI-powered voice assist. From Cortana to Siri to Google Assistant (Who needs a human-sounding name?) to Alexa, devices are getting smarter.
Richard Shapiro, who I quoted last year in Help Desk Technology Trends for 2016, included chatbots in his 2017 customer service trends piece. “More and more, chatbots are being employed to customize the user experience,” Shapiro writes. “Today’s chatbot delivers a direct, easy channel for communicating with a brand while making logical guesses about what a customer wants to do based on their previous actions.” Bots can tell customers information such as flight data and what’s on sale at a lower cost per contact for businesses. American Express’ customers can ask its Facebook Messenger chatbot for receipts, to track charges, and to help them redeem reward points.
Earlier in the year I advised brands against using Facebook Messenger for customer service. The reason I gave was that when you use email to deal with customer service, you own those email addresses. Meaning you can contact them for any reason at any time for free.
But with Facebook Messenger, the company owns the only way to get in touch with your customers, unless they give you their email address. And just as Facebook started charging you to reach your Facebook fans after you spent all that time and money building that audience, you can bet, if Facebook Messenger is successful, they’ll charge you to use that to reach them as well.
However, the chatbot function may have me revise my recommendation. There are now 34,000 chatbots on Facebook Messenger, according to David Marcus, Facebook’s vice president of messaging products. After admitting that many of the early chatbots were “really bad” at their jobs, Marcus said in November that six months after their launch, chatbots were showing dramatic improvement.
“Social networks became even more powerful in 2016, with Microsoft purchasing LinkedIn and Snapchat valued at $25 billion,” Forbes’ Blake Morgan wrote. “Facebook made a huge customer service push with the launch of Facebook chatbots offering customers the ability to make simple transactions. We’re all waiting for Snapchat’s API to configure customer service applications.”
“With Facebook’s big push into messenger bots, 2016 is being called ‘the year of the chatbots,’” or so says Help Scout’s Mat Patterson, host of the new monthly customer service industry news and trends web series called The Supportive. In episode one, Patterson discusses chatbots.
Patterson references Casey Newton’s haunting Verge story about a woman who “resurrected” her best friend using artificial intelligence.
In the comments on Patterson’s post, Anastacia Brice writes about how she loves her appointment setting chatbot, “But I also know that when bigger companies (like Verizon) want to hide their people behind bots, it just pisses me off, royally.” Brice wrote. “If I want a person, I want a person, and not easily letting me get to one is more likely to make me walk away than anything. My experience there is that, for me, they’re too robotic, and I find them frustrating.”
The best use cases for chatbots on Facebook Messenger? Driving people toward subscriptions, helping people make small purchases, and providing customer service, according to tech reporter Chris O’Brien writing in VentureBeat.
Here are some of my suggestions, with examples cribbed mostly from Slack’s Head of Developer Relations Amir Shevat writing for VentureBeat.
Use case: Bots for understanding your customers
It’s always a pain to try to survey your customers. Why not gamify it? The Swelly bot gives users a “this or that” game to learn their preferences. Click an option, and then see what everyone else chose.
Use case: Chatbots for upselling
If you’re paying attention, you can glean a lot of personal information about someone from a conversation. Because bots converse with users, they tend to gather a lot of data about them. Smart companies are using this data to better personalize ad experiences and improve click-through rates.
Andy Mauro, cofounder and CEO at Automat, built a teen influencer bot that offers coupons to start a conversation. Sentiment analysis showed 88% of users who interacted with the ad-serving bot said they loved or liked it.
Serving relevant, well-targeted coupons at just the right time is a perfect way for chatbots to upsell existing customers. Who wouldn’t want a coupon for delivery food service to arrive around when you usually eat dinner while you’re playing your favorite game (and therefore obviously haven’t been cooking)? Or the bot could offer you an easy way to buy virtual goods like tokens in a game.
Use case: Chatbots for subscriptions
VentureBeat reports that subscriptions are the most common way that bots bring in cash money for brands. This makes sense as subscription services tend to enjoy higher average customer lifetime values than companies using other payment models.
Growbot is one bot aimed at growing your subscriber base. It scans Slack conversations for keywords like “good job” or “kudos.” It then amplifies the messages by responding to them with more congratulations, promoting collaboration, rewarding good work and helping to foster positive feelings among team members.
Growbot’s premium service includes company data analytics dashboards, perfect for big companies and HR departments.
The chatbot hater brigade
Not everyone is completely on board the bot train. “If you actually talk to consumers, the people supposedly buying from all these new bots, you’ll quickly find that they actually like talking to people most of the time,” Paul Adams wrote. “I believe we are at least a decade away, if not much more, from computers being able to accurately interpret human emotion in communication.”
The problem, according to Adams, is that companies are using bots where cards would work better. They’re “blindly applying bots to interaction design problems that already have better solutions.”
“People like talking to people,” Adams wrote. “And if there is anything broken about customer communication today, it is that business online is impersonal. Businesses hide behind bad technology, like do-not-reply email, ticket numbers, automated replies, and faceless ‘contact us’ forms.” Amen.
Mat Patterson recommends following Ellen Huet at Bloomberg Technology, who’s doing awesome reporting, including The Humans Hiding Behind the Chatbots and Pushing the Boundaries of AI to Talk to the Dead.
I don’t expect chatbots to be a must-have for customer service in 2017. But I do think they will grow in usefulness and popularity. As they get used more, their algorithms will improve. Keep up to date with the topic, and when the price gets low enough and the technology good enough, plan to run some experiments.
Have you used chatbots before? What was your experience? Let me know in the comments.
Looking for Customer Service software? Check out Capterra's list of the best Customer Service software solutions.