Chatbots are becoming ubiquitous. They help us shop, tell us the latest news and weather, schedule meetings, book hotel rooms, provide customer service, and coach or tutor learners.
Wait. Coach or tutor learners? That’s right; chatbots can be found all over eLearning. See “A Mobile Coach Can Help eLearning Stick” for an example.
A chatbot, according to Mobile Coach CEO Vince Han, is a computer system—with a personality—that can automate a conversation through an electronic means, such as text, chat, Facebook Messenger, or the like.
The reason that chatbots are popping up everywhere is simple: People are increasingly using text services, even as they spend less time on email and phone calls and are less inclined to download yet another app. A key benefit of texting is that nearly all cellphones can do it, even the increasingly rare non-smartphones. Texting does not require a specific app, and nearly everyone with a cellphone does it, at least occasionally.
Pew Research Center’s “U.S. Smartphone Use in 2015” study found that virtually all smartphone users aged 18 to 49—and 92 percent of smartphone users aged 50-plus—used texting at least once during the week of the survey (see References at the end of this article). Pew also found that “text messaging is a global phenomenon—across the 21 countries surveyed [in 2011], a median of 75 percent of cellphone owners say they text.” And, according to Gallup, “sending and receiving text messages is the most prevalent form of communication for Americans younger than 50” (see References).
In short: Chatbots are an easy way for people to interact with companies and services using a technology they are comfortable with and already use. In a recent webinar on artificial intelligence and corporate training, Han sketched out the difference between ordering a pizza using an online order form versus using a chatbot.
To use the online form on your smartphone, Han said, you’d need to:
- Download the app
- Create an account
- Enter your address
- Explore coupons, offers, and possibly several pages of menu choices
- Navigate multiple options to make your selections
- Check out and pay, including entering payment information
In contrast, using a chatbot would be simple and conversational—and wouldn’t require that the customer download an app for each new service used. Texting with a chatbot fits easily into the flow of most employees’ daily lives.
A simple chatbot uses preprogrammed responses and follows intelligent rules or heuristics to select the appropriate response. These chatbots can run into problems if the human says something that the chatbot programmer did not anticipate. A more sophisticated chatbot uses machine learning to figure out responses to whatever the human enters, based on previous conversations. The most sophisticated chatbots generate responses, rather than using a library of stored response. These can reference information from previous texts, and they constantly improve their ability to respond to “everyday” language. But this chatbot is more likely to make grammatical errors—and it requires considerable “training” before it’s ready to chat.
So, how might any of these chatbots be useful in eLearning?
Mobile Coach’s answer is coaching and follow-up to ensure that training sticks. But there are lots of other possibilities:
- Enrollment—a chatbot can be “taught” the prerequisites and other requirements and then enroll eligible learners in the correct courses, saving human staff a lot of time.
- Simulations—the chatbot can take the part of any participant in an interactive simulation, bringing eLearning activities into messaging and engaging learners in realistic practice conversations.
- Answering learner questions—the chatbot can be “taught” common questions and respond immediately to learners’ questions.
- Quizzing learners—chatbots can quiz learners on vocabulary or other fact-based learning to prepare for quizzes, ensure that learning sticks, or just for fun. An intelligent chatbot can even adapt, personalizing the questions asked or information reviewed to the individual learner, and adjusting to the learner’s responses.
- Assessment—chatbots can administer quizzes or other assessments and collect responses.
- Onboarding—hiring a new learner can trigger the chatbot. According to Han, activating it can be part of the hiring process; the chatbot then can walk the new learner through required processes, provide links to forms and prompt the learner to complete them, provide an office tour, supply and remind the learner of benefits information and registration deadlines, and even suggest popular lunch spots near the office.
If it sounds too good to be true, some of it still is. The machine-learning chatbots are still in early days; in many cases, it is obvious that the learner is interacting with a chatbot, not a human. But so what?
While good chatbots that can teach leadership skills, or help employees practice difficult conversations, might not be here yet, chatbots can and do teach learners about new product features, guide new employees through onboarding, and teach simple processes and procedures. People have proven ready and willing to chat with the chatbots, and, Han says, they work.
Mobile Coach’s study of learners in a pharmaceutical industry training course found that 95 percent of the learners reported that the chatbot coach was helpful; nearly all the learners rated the course better than learners in previous years, sans chatbot, had rated it. A different study found that learners with a chatbot coach were 38 percent more likely to follow through on key commitments than learners in a control group who had no chatbot to remind and prompt them.
Newport, Frank. “The New Era of Communication Among Americans.” Gallup. 10 November 2014.
Pew Research Center. “Global Digital Communication: Texting, Social Networking Popular Worldwide.” 29 February 2012.Smith, Aaron. “U.S. Smartphone Use in 2015.” Pew Research Center. 1 April 2015.