What we learned from our AI study.
We found that AI is an effective tool to help people quickly develop and improve their leadership coaching skills, but there are some key details for companies to consider.
Our data shows that AI, when used effectively as a tool to augment training activities, can be an opportunity to deliver learning and development activities “at half the time, half the cost, with twice the impact.”
Our study involved 18 managers across a wide range of industries, each with an average of 12 years of leadership experience. Our question was “to what extent can AI impact soft-skills learning and application, particularly leadership coaching skills?”
58%
Increased confidence in using a coaching model with employees
Summary of key data
187%
Increased comfort talking to AI versus typing into the bot
9.2
Rating for how helpful AI was viewed as a learning tool
9.4
Rating for how helpful AI would be to help prepare for coaching conversations
“The AI bot gave practical feedback that can be implemented immediately.”
We used a free, publicly available AI chatbot and asked participants to answer pre- and post-study questions. Participants completed seven modules focused on coaching skills and the well-known GROW coaching model (Goal, Reality, Options, and Way Forward), using prewritten prompts to initiate conversations with the bot. There was no human intervention before or during the study. All participants used the same AI bot and spent an average of 60 minutes completing the seven modules.
Key findings
Our study shows AI can be a critical piece of a soft skills learning and development effort. It can provide extreme customization unachievable in human-only-led workshops or traditional self-paced learning.
AI can dramatically reduce the time and costs associated with learning and development.
Participants made it clear that they want AI to be part of the learning and development process, not a replacement for live, interactive discussions and collaboration.
The particular platform is not as important as ensuring people are comfortable with the technology. While research shows talking to the bot provides greater benefits than typing, text is still powerful.
Using AI to teach soft skills can be a lever to drive AI adoption across the enterprise and promote organizational behavior change.
“When I first thought about a bot coaching me, I thought that may not be as effective as in person. [After doing this] I would beg to differ.”
Additional insights
Participants reported significant learning in a very short period of time, despite each having over 12 years (on average) of leadership experience. This greatly surprised us.
Participants repeatedly cited AI’s level of customization and its ability to focus on any scenario as two of the most significant benefits. Some noted that the machine had “more knowledge than their manager or colleagues could provide.”
Although participants were “amazed” by how AI guided them through the process, EVERYONE indicated that augmenting the process with a human trainer, as part of a live workshop and/or an ongoing program where they could interact with colleagues, would be ideal.
Participants said completing the self-assessments led to high levels of customization (DiSC, conflict (Thomas-Kilmann), and emotional intelligence). Some indicated a desire to dig deeper into these topics.
There were sporadic issues with speaking to the bot rather than typing, and a few instances of not being able to continue a session in private browser mode. These are easily rectified but technology issues must be addressed to prevent participants from becoming frustrated or discouraged.
Participants felt that the pre-scripted prompts at the start of each module helped provide better responses from the AI bot. One participant noted, “The hands-on nature and the mix of written prompts, but room to include your own personal details (styles) and AI convos made it a very impactful experience.”
Participants noted that pace and frequency can promote habit building and behavior change, helping people become more comfortable using AI in their day-to-day jobs. People had varied opinions on how the learning should be structured. Some preferred to complete all seven modules in one sitting, while others indicated they would prefer to “have a module a day.”