Every Manager Is Becoming a Coach. Most of Them Don't Have One.

The job is changing faster than the support

In November 2025, the McKinsey Global Institute published Agents, robots, and us: Skill partnerships in the age of AI. Its headline estimate: technologies available today could, in theory, automate activities accounting for about 57% of US work hours. McKinsey is clear that this measures technical potential, not a forecast of job losses.

The more useful finding for anyone in people development is what happens to skills. More than 70% of the skills employers seek today apply to both automatable and non-automatable work. The skills aren't disappearing. How people use them is changing. Demand for AI fluency in US job postings has grown sevenfold in two years, faster than any other skill.

For managers, the shift is sharper. McKinsey describes managerial work shifting from supervising people to orchestrating teams in which people and AI agents work together. That frees managers for higher-value work such as influencing and mentorship. The report's own example: a sales manager who spends more time coaching the team to use AI-driven insights. Managers are now expected to help their people interpret AI output, exercise judgment, and work through uncertainty.

In other words, managers are being asked to coach.

Who coaches the coach?

Here's the gap. In most organizations, one-to-one coaching goes to executives. First-time managers and team leads get a workshop, a framework, and a slide deck. Yet they are the ones now expected to coach.

Picture a newly promoted manager with a difficult feedback conversation on Thursday. The feedback workshop was in March. She remembers the model in outline, but not how to apply it to this person, this situation, this week.

The problem isn't knowledge. It's continuity. Development happens in moments (a workshop, a coaching session), and the situations that need it happen in between.

Why a generic AI assistant isn't the answer

That manager can already open a general-purpose AI assistant and ask for advice. She'll get something plausible. But it won't know the feedback model her company trained her on, its leadership principles, or the values behind them. At worst, it contradicts the program the organization just paid for.

AI coaching itself isn't a fringe idea. In its own case studies, McKinsey sketches "coaching agents" that give sales and service reps personalized, real-time feedback. The real question is what those agents are grounded in.

Coaching that sticks has to be grounded in the methodology people already learned. That's why CoachBot builds every coaching agent around a specific methodology, not a generic chatbot with a logo on it.

It's also why we build a custom evaluation framework with each partner. Together we define what "good" coaching looks like for their approach, and we test against that definition before the agent reaches a single user.

Trust decides whether managers use it

A manager will only bring a real problem to a coaching tool if she trusts where the conversation goes. HR needs the same assurance before rolling anything out. Privacy infrastructure rarely wins a deal on its own, but it is often the reason one stalls.

CoachBot is GDPR-compliant and EU-hosted, and an ICF Germany Charter Chapter Business Partner. Coaching quality is overseen by our Responsible AI Coaching Quality Council, chaired by Chief Coaching Officer Jonathan Reitz.

We don't train models on sensitive client information. In fact, we don't fine-tune any models today.

Organizations see aggregated, anonymized results rather than individual conversations.

CoachBot is currently in its SOC 2 Type II audit, with the report expected by the end of Q1 2027.

Why this matters

• For coaching providers: If every manager is expected to coach, the demand for your methodology just grew far beyond what one-to-one delivery can serve. Encoding your approach lets it reach managers between sessions and at scale. Your coaches can then spend their time where human depth matters most.

• For HR leaders: You no longer have to choose between real coaching for a few and a workshop for everyone else. You can extend coaching down the hierarchy, grounded in the frameworks you already invested in. Human coaching stays where it adds the most value.

AI is changing what managers do. The open question is whether anyone is helping them do it.

Organizations keep asking their managers to coach their teams through AI transformation. Most of those managers don't have access to a coach themselves, and some have never had one. Closing that gap, without replacing the human coach who makes coaching so valuable in the first place, is what we're building for.

Lewin Keller, Founder and CEO, CoachBot AI

If you're exploring how this could fit your development strategy, a focused pilot is a practical place to start. Reach us at team@coachbot.ai.

Sources

• McKinsey Global Institute (November 25, 2025). Agents, robots, and us: Skill partnerships in the age of AI.

Related reading

• CoachBot. Why Building Your Own AI Coach Is the Easy Part.

• CoachBot. Why Enterprise Coaching Investments Aren't Delivering — And What L&D Leaders Can Do About It.

Lewin Keller

Founder & CEO, CoachBot AI, ICF-trained coach (ACTP)