Coaching works, but it’s expensive. You want to uplevel your first-line managers, and you would ideally love to provide 1-1 coaching to all of them—to every employee, if you could—but your budget only covers a small fraction of that.
The solution? The AI coach.
In the coming years, AI will drive seismic change in coaching. It will augment human coaches and replace much of their current work. It will make high-quality coaching available to a vastly larger marketplace. At senior executive levels, there will be an ecosystem of human coaches, AI coaches and vital partnerships between them.
For HR leaders who oversee coaching programs, this raises three key questions:
- What can AI coaches do for my company now?
- How should I implement them?
- And, what will AI coaching look like in the future?
Today’s AI coaches offer on-demand, company-tailored support for performance, goal setting, role-play and behavioral nudges via text or voice. Some are beginning to use “digital humans”—human-like avatars with faces and voices you can interact with as if they were real people. AI coaches are being tested and scaled at Fortune 500 companies.
One example, Valence, reports promising results with its coach “Nadia.” In one case, 90% of participants use Nadia regularly with a net promoter score of 91. BetterUp is piloting “Grow,” an AI coach that boosted workplace confidence for 16% of survey respondents, with 95% reporting overall satisfaction. CoachHub’s AIMY is another early mover in this space, which also offers human coaches. Companies like Virtual Sapiens and Yoodli are focused on AI role-playing tools to improve communication skills in roles like sales or public speaking.
How to implement AI coaching
When thinking about how to implement AI coaching at your company, design your program based on the respective strengths of AI and human coaches. Here is a simple triage model.
- Use AI coaches for access, skill building and data.
Access: AI makes coaching available to everyone in your organization, 24/7. AI coaches are already getting good at the basics: listening, asking useful questions and mimicking empathy. They can role play, provide guidance and nudge clients to practice new behaviors.
Skill building: AI coaches can help employees build professional and managerial skills. Employees who need help getting organized, managing their time, communicating with colleagues, delivering performance feedback or improving their public speaking will find useful guidance and practice with an AI coach.
Data: An AI coach can process huge volumes of information, drawing on organizational data, leadership models and company values. Because coaching is so personal and trust so crucial, it’s important to verify that platforms you’re considering meet privacy and security standards such as GDPR and SOC 2 Type II certification.
AI coaches have access to a vast universe of knowledge. So, an employee could talk with their AI coach about a challenging direct report one moment, and then pivot to a deeply technical topic the next. Remember, though: AI coaches do hallucinate at times, presenting counterfactual information with great confidence.
- Use human coaches for high stakes, high complexity and high touch.
High stakes: While an AI coach might be helpful in talking through day-to-day business challenges, human coaches bring lived real-world experience, context and intuition that make them more reliable guides in mission-critical situations; for example, as a sounding board when a senior executive is facing a crisis or preparing congressional testimony.
High complexity: Senior executives manage large, complex global organizations, and have complex relationships with board members, their CEO and peers. Human coaches can more effectively engage with all of this. They can also communicate with the stakeholders invested in the leader’s development (manager, HR, L&D, other consultants) and manage nuanced confidentiality with each of them.
High touch: Even as people become accustomed to digital humans, most of us will recognize that they are not real people. Human coaches offer real relationships that enable transformational growth. They are experts in human emotions, psychology and relationships.
Importantly, humans have bodies, so they can be present, meet in person, go for a walk. This opens the fullest possible throughput of communication between coach and client, including all the subtleties of voice, body language and shared mood.
- Use a hybrid approach for situations in the middle, or to get the best of both.
Using an AI coach or a human coach is not a binary choice: One employee can use both.
Situations in the middle: Many first-line managers and middle managers need a combination of skill-building and occasional higher-touch, higher-complexity work. These are prime candidates for a hybrid approach, using an AI coach as much as they want, with occasional access to a human coach.
Getting the best of both: Some senior leaders whose roles demand the capabilities of a human coach will also benefit from an AI coach because of their constant accessibility and data superpowers.
What’s next: the evolution of AI coaching
An AI executive recently said to me, “However long you think something will take to happen in AI, it’s going to happen much faster.” The velocity of AI is rapid, so AI coaches will improve with surprising speed.
Soon, AI coaching will be available to employees at most companies. It will boost performance and become a crucial tool in the L&D toolkit. In a large company, thousands of employees might interact daily with an AI coach—and that coach will be an ambassador to drive the culture the company most needs.
Human coaches will continue to provide unique value. Much of what they do today for frontline employees and managers will be done by AI coaches, but for a narrower set of needs at those levels, and for CEOs and senior executives, human coaches will still be essential.
HR leaders should stay informed as new capabilities emerge. AI coaches will do more and more of what human coaches do today, so coaching strategy will need to evolve each year.
Credit: Source link
