Why the AI Coach That Challenges You Hardest Earns Your Trust
For decades, coaching research has pointed to the same conclusion: it doesn't matter which model you use. GROW, Solution-Focused, Person-Centred, Insight-Oriented — the outcomes are statistically equivalent. What matters is the relationship between coach and client. The methodology is almost incidental.
This finding, known as the Dodo Bird Verdict, has shaped how the entire coaching profession thinks about its own effectiveness. Your chosen framework doesn't predict your impact. Your presence, your congruence, and the quality of the working alliance does. For human coaches, this is both reassuring and — if you sit with it — quietly unsettling.
Then someone asked the obvious next question: what happens when you remove the human entirely?
The Study That Changes the Conversation
In 2024/25, researchers Rebecca Rutschmann, Prof. Nicky Terblanche, and Jonathan Reitz ran a two-wave experimental study with 592 participants comparing three AI coaching chatbots: one using the GROW model, one Solution-Focused, and one grounded in Cognitive Behavioural Coaching (CBC). They measured impact on goal attainment, working alliance, and technology adoption.
The assumptions going in were reasonable. The Solution-Focused bot would probably lead — it's warm, affirming, future-oriented. The CBC bot might struggle — cognitive behavioural coaching is demanding, and that might feel heavy through a screen. GROW might feel too procedural.
The results overturned every expectation.
The CBC coachbot outperformed both alternatives on working alliance and technology adoption — including whether users enjoyed it and intended to use it again. The most cognitively demanding approach was experienced as the most enjoyable and the least effortful.
Goal attainment was equal across all three bots — any well-designed AI coach can help you reach your goal. Where CBC won was the quality of the experience: the sense of alliance, the engagement, and critically, the motivation to come back.
When the Human Leaves, the Architecture Speaks
The Dodo Bird Verdict rests on the idea that common factors — presence, warmth, trust, a credible framework — do the heavy lifting. Strip those away and you're left with pure architecture. And when pure architecture was measured, one approach won clearly.
Why CBC? The hypothesis the researchers find most compelling is that CBC's Socratic questioning does something emotionally and cognitively specific. It doesn't affirm. It challenges. It invites you to examine your own thinking rather than guiding you toward a predetermined positive frame. In doing so, it creates what transformative learning theorists call reflective dissonance — the productive friction that precedes genuine change.
There's also a timing dimension worth noting. We are living through one of the most uncertain periods most leaders have navigated. CBT was originally developed to treat depression, but it built its strongest evidence base in treating anxiety disorders — conditions defined by exactly the kind of looping, catastrophic thinking that structured cognitive challenge is designed to interrupt.
The logical, step-by-step progression of CBC — the sense that there is a way through, that your thoughts can be examined and reorganised — may offer something particularly resonant right now. Not escape from discomfort, but a map for moving through it.
What This Means If You Lead Under Pressure
Most coaching — human and AI — defaults to warmth. It affirms. It validates. It creates psychological safety by minimising friction. And there's nothing wrong with that, except when it means avoiding the question you actually need asked.
This study suggests that trust isn't built by making you comfortable. It's built by challenging you with enough structure that the challenge feels safe. Psychological safety doesn't mean the absence of friction. It means the presence of enough trust to tolerate friction.
AI, which has no ego investment in being liked, builds trust through challenge rather than warmth. It has no political stake in your decision. It doesn't need you to leave feeling good. It needs you to leave thinking clearly.
For senior leaders, this is the gap that matters. Your direct reports filter bad news. Your board wants confidence. Your peers are competing. Everyone around you has an agenda. The coaching that tells you what you want to hear fits the pattern perfectly — and changes nothing.
How The Deep Brief Is Built on This Principle
The Deep Brief was designed as a challenge-based coaching system before this research was published. The methodology comes from a different source — military command decision-making and deep technical diving, where comfort kills and structured cognitive discipline keeps you alive. But the research validates the design.
Every session ends with a named decision, not an insight. The coach tracks your commitments and opens the next session by asking whether you followed through. It spots patterns across conversations — the things you circle but don't name — and puts them in front of you. It doesn't affirm your thinking. It pressure-tests it.
The research says this approach builds stronger trust, higher engagement, and greater return intention than the alternatives. The experience of leaders using The Deep Brief confirms it: the sessions that feel hardest are the ones that change something.
The Research
Rutschmann, R., Terblanche, N., & Reitz, J. (2024/25). Comparing GROW, Solution-Focused and CBT Coaching Chatbots: A Self-Determination and Transformative Learning Perspective. Two-wave experimental study, 592 participants. Read the original article →
De Haan, E. How Can Coaches Choose Their Approach and Their Interventions Based on the Evidence We Now Have. Referenced for the Dodo Bird Verdict framework and human coaching outcome equivalence.