By Bonnie Beresford

A chatbot for learning? The answer to that question is more and more commonly becoming “yes!” But what is the attraction and how do we know if they really work?

Organizations invest millions of dollars in learning and development (L&D) yet are often disappointed when performance doesn’t improve as expected. Why? Even great training won’t deliver results without reinforcement. Research tells us the importance of spacing and repetition in mastering new knowledge and skills. It also tells us the importance of the manager in learning transfer. So, what is happening?

While L&D organizations try to engage managers to help employees apply learnings after training, this engagement doesn’t happen often enough. Most managers are ill-prepared—or simply don’t have the time—to support and reinforce their people on the specifics of the training. As a result, learning transfer is often not happening.

What if L&D teams could help close this gap? What if we could provide the timing, spacing, and relevant reinforcement without relying on the manager? New technologies give us an opportunity to take a fresh look at the learning transfer problem. In fact, it is a perfect use case for a chatbot.

Today, most people have smartphones, and many have unlimited data plans. By using existing technology and easy-to-use scripting software, a chatbot can become the coach L&D has always wanted. Bots can readily extend the learning and provide reinforcement. Chatbots are popping up across the learning spectrum, from acronym databases for new hires to competitive comparison information for sales teams. We are also starting to build case studies that seek to show if chatbots work and under what conditions.

A word of caution though, when jumping into a new learning technology—just because the technology works does not mean your application will do the same. Deployment of a chatbot, as with any new technology implementation, requires addressing three factors: the technology, the content, and the change management.

Like a three-legged stool, if any one of these factors is missing, the implementation will likely be substandard, if not a complete failure.

The Technology: The best way to know if the technology works is to play with it. Most vendors will provide access to a sandbox for prospective clients to test out the tool. Build out a small section of your content then run a pilot. Don’t forget to check out the stability of the vendor to be sure they will be around down the road. In all my research on learning technology deployment, the technology itself is rarely the weak leg of the stool.

The Content: You want your content to grab your users and provide the reason they should continue to engage. Start small, with impactful content tailored to your audience. One implementation of a learner experience platform started with more than 700 unique pieces of content; after the first month, more than 500 pieces had never been viewed, while 20 pieces had 80 percent of the hits. Again, a pilot can help home in on the content that matters most.

The Change Management: This area seems to pose the greatest challenge and where most deployments fail—getting learners to change their behavior and use the new technology. How will you know if it’s working? How will you measure engagement and adoption? You need to get beyond numbers of learners enrolled and look at the frequency and duration of engagement as well as high-usage content. Perhaps most important for getting a new technology to take hold is discovering under what conditions learners most readily adopt it. Is it when the technology is introduced and demonstrated in a live training session? Is it if it is required? Does manager buy-in matter? Running a pilot using various approaches then comparing engagement and adoption rates can help you figure out what works best for your audience.

Chatbots, like other new learning technologies, can play a meaningful role in your learning ecosystem. Before going all-in, design a pilot and define how you will measure its success across all the legs of that three-legged stool. Then track adoption and get learner feedback so you can leverage these new technologies and deliver an extended learning experience.

This article originally appeared at