Can AI Replace a Business Coach? What I Learned From Around 2,500 Sessions

I have spent around thirty years coaching entrepreneurs, founders, leaders and management teams. From young companies still discovering what they wanted to become to international brands.

Then I tried to teach AI how I coach.

Not just the questions I ask, but also how I probe deeper, connect previous answers, recognise contradictions and offer empathetic suggestions when someone is struggling to get to the heart of the matter.

That led to the Business Strategy Canvas Builder: an AI business coach inside ChatGPT that guides people through a complete strategic process.

I then analysed around 2,500 registered user sessions to understand what actually happens when people work independently on their business with an AI coach.

My most important conclusion?

AI can already perform far more of the work of a good business coach than most people realise. But even the best question has little value if someone is unwilling to stay with the answer.

That is not a technical problem. It is a human one.

I Did Not Build a Chatbot With a Questionnaire

The Business Strategy Canvas Builder consists of eleven connected steps:

1. Validation & Business Name

2. Dream

3. Purpose

4. Big Why

5. Role

6. Entity

7. Strategy

8. Target Group

9. Products & Services

10. Rules of the Game

11. Presentation

Each step builds on what has been said before. At the end, the user can generate a complete strategic presentation.

But the intelligence is not in those eleven steps.

A conventional questionnaire collects answers. This AI business coach investigates whether those answers genuinely fit together.

When someone gives a generic answer, the AI probes further. When the chosen strategy does not align with the Purpose defined earlier, it points that out. When two answers contradict each other, it brings that tension back into the conversation. And when someone gets stuck, it can offer suggestions based on everything they have already shared.

I did not simply put my questions into ChatGPT. I tried to make my way of observing, listening, analysing and coaching transferable.

That is a fundamental difference.

Thirty Years of Coaching Translated Into AI

The methodology behind the Business Strategy Canvas Builder was not created at a desk.

Since founding my first company, Quince, in 1994, I have built businesses and coached entrepreneurs, founders, leaders and management teams. Later, I continued that work through WeMindd and bensteenstra.com.

I have worked and coached within startups, established companies and international organisations. Not just on communication, marketing or positioning, but primarily on the questions that come before them.

What do we really want to achieve?

Why does this company exist?

What role do we want to play?

What do we believe in?

Which choices still fit, and which no longer do?

I incorporated that experience into the AI. As a result, the Business Strategy Canvas Builder does not merely know which strategic subjects need to be addressed. It also understands how the answers should relate to one another.

Sometimes the AI asks exactly the question I would have asked. Sometimes it even finds an angle sharper than the one I would have thought of at that moment.

That does not make it better than a good, experienced coach. It mainly shows how much professional knowledge and coaching methodology can be transferred when you learn to recognise and structure it carefully enough.

An AI Business Coach With a Reach I Could Never Achieve Alone

The biggest difference is not just speed, availability or price.

The Business Strategy Canvas Builder operates inside ChatGPT. This means my methodology can be used in more than seventy languages by people from over ninety countries.

As an individual coach, I could never achieve that reach.

An entrepreneur does not need to travel to the Netherlands, speak the same language as I do or arrange a personal appointment with me first. Someone can begin the moment a question becomes urgent.

That makes high-quality strategic reflection accessible to people who would normally never sit across the table from me.

But accessibility is not the same as commitment.

That became clear in the research.

What Around 2,500 Registered Sessions Reveal

For this research, I analysed a defined dataset of around 2,500 registered user sessions within the Business Strategy Canvas Builder.

These do not represent 2,500 unique people. Nor do they represent every coaching conversation or digital session that has ever taken place. This is a specific research dataset consisting of registered sessions.

Around 17 percent of the registrations involve tests, duplicate sessions or restarted processes. These have not yet been removed from the raw dataset.

Within that dataset:

• Almost 600 sessions reached and confirmed the second step, Dream.

• 130 sessions reached and confirmed the third step, Purpose.

• 34 sessions completed the entire process and generated a strategic presentation.

• Around 0.8 percent stopped because of a recorded technical problem.

Those 34 completed presentations demonstrate something important: with the help of AI, someone can independently move from an initial business idea or strategic problem to a coherent strategy and final presentation.

Without me being present in the conversation.

But the other sessions are at least as interesting. They reveal what people do when an intensive strategic process becomes freely and instantly available.

Some people wanted to explore a single question. Others were mainly curious. Some began seriously but stopped when the thinking became more difficult or personal.

An unfinished session is therefore not automatically a failed session. One good question can already set something in motion. But creating a complete and consistent business strategy requires the different elements to be connected.

The AI Was Rarely the Technical Problem

The drop-off was not primarily caused by malfunctions, delays or unclear interactions.

These elements are monitored. Recorded technical problems account for around 0.8 percent of abandoned sessions.

Time does not appear to explain everything either. After excluding extreme interruptions, the measured time per completed step ranged from around 54 seconds to 12 minutes. The time someone needs differs greatly by person and question.

The pattern also remained reasonably stable across different weeks and months. It therefore does not appear to have been a random peak or temporary technical effect.

Users cannot currently resume a process they have already started. Someone who begins again does so in a new session. This may explain some of the recorded restarts, but not the entire pattern.

The interesting question is therefore not simply why the AI failed to move someone forward.

The more interesting question is why someone decided not to continue.

What Users Said About Stopping

Around 7 percent of the registered sessions responded to an exit question that I later sent by email. This is a limited group, so the answers cannot automatically be applied to all sessions.

Within this group, around 23 percent selected the answer:

“I just wanted to have a look.”

Around 67 percent of this group already had an existing business.

A separate 15 percent said they only wanted to understand how the Business Strategy Canvas Builder worked and had no interest in creating a business plan.

This confirms that a free AI tool also attracts people who never intended to complete an entire strategic process.

But curiosity does not explain everything.

The transition from Dream to Purpose stands out in particular.

Why Purpose Appears to Be the Turning Point

Dream is attractive.

What do you want to build? What do you dream about? What would success look like if you did not immediately limit yourself?

Questions like these often generate energy.

Purpose asks something different.

Why should this company exist? What do you truly want to change? Why is this important enough to deserve years of your life, attention and responsibility?

That is not a marketing statement. The answer can have consequences.

Of all the sessions analysed, around 40 percent involved a new or not-yet-established business. Approximately 60 percent involved an existing business.

Among the sessions that reached Purpose, that ratio was almost reversed:

• Around 72 percent involved a new business.

• Around 28 percent involved an existing business.

Because these percentages come from the same population and research setup, they suggest that new entrepreneurs were considerably more likely to reach Purpose than existing entrepreneurs.

The data does not yet prove why.

My professional experience does suggest a possible explanation.

A new entrepreneur is often still actively connected to the original ideal. Everything is open. The company still largely consists of possibilities, beliefs and ambitions.

An existing business is different. Over time, the original idea becomes surrounded by customers, employees, revenue targets, processes, obligations and compromises.

The entrepreneur does not arrive asking:

“What is my Purpose?”

They arrive because their motivation has disappeared. Because revenue is disappointing. Because the team is no longer functioning properly. Because growth has stalled. Because the positioning no longer fits. Or because they no longer know whether they want to continue.

At that moment, Purpose may seem too abstract.

But that may be exactly where the real problem lies.

Not in the next campaign, a new positioning strategy or yet another growth plan, but in the lost connection with what once made the company worth building.

The Confrontation You Can Easily Leave Online

An AI business coach can ask a sharp question. It can probe deeper and connect different ideas. It can even reveal that the strategy someone claims to want does not fit the company they have actually built.

But the user can also close the conversation immediately.

That is the freedom of a digital coach. It is also its weakness.

There is no scheduled appointment. No team remaining at the table. No other person noticing that you are avoiding the real issue. No shared decision that this question deserves everyone’s full attention today.

That is an important difference between the AI and my facilitated strategy sessions.

Such a session usually lasts between four and eight hours. Depending on the problem, the number of participants and the complexity involved, the investment ranges from €3,000 to €9,000.

Across the different forms in which I have conducted these sessions over the past thirty years, more than 200 and possibly closer to 250 times, the process is completed. The participants then actively use the outcome.

More than 90 percent of these sessions involve existing entrepreneurs or organisations facing a specific problem. Sometimes it is financial, sometimes cultural and often related to direction, motivation or coherence.

That does not mean the price causes people to complete the process. Nor do I know which element is decisive. The appointment, the reserved time, my questions, the confrontation, the presence of other participants and the guidance may all play a part.

This is not a controlled experiment.

But the contrast matters.

The questions are available online. The methodology is available. The AI can guide the process.

Yet access to good coaching is clearly not the same as being willing to let yourself be coached.

AI Coach Versus Business Coach

Can AI replace a business coach?

My answer is: AI can take over a substantial part of the work of a business coach. For some people, AI can even guide the entire strategic coaching process.

The 34 generated final presentations show that this is not a theoretical possibility. It is already happening.

AI can ask questions, probe deeper, analyse, formulate, connect previous answers and reveal inconsistencies. It can also do this immediately, internationally and in more than seventy languages.

A good human coach adds something different.

Not because that coach should necessarily be smarter, more attentive or more empathetic than AI. An experienced coach is equally capable of being fully present, remaining sharp and remembering what truly matters.

The added value lies in the human context.

In what happens between the words. In the dynamics between entrepreneurs, partners or management team members. In sensing where tension arises. In slowing someone down when they produce a rational answer too quickly. And in holding everyone’s attention when an insight becomes uncomfortable.

AI can learn a significant part of how I coach.

But no coach, human or digital, can take responsibility for someone else’s honest answer.

The Real Lesson From This Research

The easiest story would be that people stop because the AI is not yet good enough.

My research does not automatically point in that direction.

The AI can guide the entire process. Technology explains only a small part of the drop-off. The methodology works across different languages and cultures. Users can arrive at a complete strategy and presentation.

The more difficult conclusion is that access to quality does not automatically mean that people will use that quality.

We often ask whether AI can become intelligent, empathetic and human enough to replace a good coach.

Perhaps that is no longer the most important question.

The more important question is what someone needs to stay with a difficult subject. To avoid stopping at the first sensible answer. To recognise that a visible business problem may begin with something they would rather not examine.

That is also the next challenge for AI coaching.

Not simply adding more questions, but becoming better at recognising what someone needs to continue. Sometimes that means a clearer explanation. Sometimes a different formulation. Sometimes a smaller step. And sometimes it means reaching the honest conclusion that the conversation would be better conducted with a human coach, business partner or management team.

My Conclusion About the Future of AI Coaching

AI will replace part of traditional business coaching.

I do not see that as a threat. It means that good strategic guidance becomes accessible to people, languages and countries I could never personally serve on my own.

At the same time, AI makes the real value of coaching more visible.

It is not about possessing a secret list of questions.

It is not about providing answers someone could have thought of themselves.

It is about creating a situation in which someone is willing to look honestly, make sharper choices and take responsibility for what they discover.

The Business Strategy Canvas Builder shows that a coaching methodology can be made scalable.

The analysis of around 2,500 registered sessions also reveals what does not scale automatically: attention, urgency, commitment and the willingness to stay with a difficult answer.

That is why my principle remains the same:

If it feels right, it works.

Not because every step should feel pleasant or easy, but because a strategy only truly works when it aligns with who you are, what you want to build and why you are willing to choose it.

About the Research

This article is based on a defined analysis of around 2,500 registered user sessions within the Business Strategy Canvas Builder.

A session ID does not necessarily represent one unique individual. Around 17 percent of the registrations consist of tests, duplicate sessions or restarts. These have not been removed from the raw research data.

The steps mentioned were actually completed and confirmed. Users cannot currently resume a process they have already started. A new attempt is registered as a new session.

The exit question was answered by around 7 percent of the sessions. The responses are therefore indicative and not representative of all users.

The comparison with my facilitated strategy sessions is based on my professional experience and not on a controlled scientific experiment. The data shows what happened. The explanations in this article are interpretations and hypotheses based on that data and approximately thirty years of experience as an entrepreneur, strategist and coach.