Mid-market CEOs I have coached in neurocoaching
These four accounts are real. To protect the leaders who trusted me, I have changed first names, company names, precise sectors and some figures. Everything else is faithful: the starting point, how the work unfolded, the hard moments, the results. The last one is my own: the 2008 crisis, lived from the inside.
If one of them resembles yours, call me: thirty minutes, no sales pitch.
Book an appointmentThe situation
In 2022, Olivier’s company turns over 6 M€. He founded it, he runs it. The potential is there; so is the pressure.
The first piece of work is not the company. It is him: his personal goals, what really matters to him.
What we did
The work moved forward on two fronts at the same time: the leader and the company.
- Refocusing his personal goals on what really matters to him, before talking about the company’s.
- Learning to manage stress. Olivier says it himself: that is the step that took him up to the next level.
- Strategy advice and structuring of the company, as a former CEO myself, session after session.
Four years on
The company has doubled: 12 M€ in revenue today. And despite the crisis, it still believes in the future and keeps growing.
- 6 → 12 M€in revenue in four years
- 4 yearsof engagement, still ongoing
The situation
Frédéric is 57. He has run this company since 2002. In March 2024, after a long process, he sold 78 % of the capital to a French mid-cap fund, keeping the chair and a minority stake. The mandate was clear: structure the company for a secondary sale in thirty-six months, lifting EBITDA from 11 % to 14 %. Frédéric knew how to run the operation: twenty-two years of experience. He did not know how to industrialise it. The fund was pushing hard on the roll-out of artificial intelligence (AI). His IT director, in post since 2009, pushed back at every committee meeting. Frédéric read Votre cerveau, fournisseur officiel de bonheur when it came out, in the spring of 2024, then met another Lille business leader I was working with. He wrote in August.
The Leader’s Brain assessment, carried out in September 2024 in his office in Lille, revealed a rare profile: excellent resistance to acute stress, but a long-masked emotional exhaustion. Frédéric was running on the company’s memory. He could no longer project himself beyond eighteen months. A quiet anhedonia had set in. The three-day AI assessment on site identified an obvious use case that nobody had dared to carry: predictive maintenance on the eighteen production lines. The average breakdown rate was 2.3 %, but 6.1 % on the six critical lines. Sensor data had been sitting in the systems since 2019.
What we did
- Fifteen individual sessions, seven of them by video call from Lille, with the Leader’s Brain assessment revisited at seven months.
- On the AI side, I first spent three months working on the relationship with the IT director, before the project was even launched.
- I coordinated the deployment of a predictive maintenance model on six pilot lines with a technical partner based in Paris, keeping the in-house IT director as internal project lead.
- In month eight, a second use case: an internal assistant that answers operators' questions by fetching the information from the company’s own documents, and citing its source (this is called RAG, for Retrieval-Augmented Generation). Deployed across the 14,000 technical sheets produced since 1998, it was meant to shorten the onboarding of new operators: eight weeks on average.
The hard moment
In April 2025, Frédéric blacked out at the wheel on a Sunday evening, on the A1, driving back from a weekend at his sister’s. No hospital stay. Six days off work. A session with me the following Tuesday, by video. He said, calmly, "maybe I started dealing with this too late". He cried for the first time over the death of his father, in 2009.
"Maybe I started dealing with this too late."
Six months on
In March 2025, the predictive maintenance pilot was running on six lines. The breakdown rate there had come back down to 3.9 %, from 6.1 % before. Gross savings for the half-year: 187 k€. The IT director had changed his attitude in February, after the Paris technical partner explicitly credited him with the design.
A year on
A year after the start, the extension to fourteen of the eighteen lines was complete. The plant’s overall breakdown rate had fallen to 2.9 %, against 4.7 % on average before the project. EBITDA for the first half of 2025 reached 12.8 %, against 11 % a year earlier, a trajectory in line with the plan presented to the fund. The document assistant has been running since April: a new operator is now onboarded in 5.2 weeks. Frédéric took up road cycling again in the spring. He agreed to a full health check in June. He also started talking with his wife about what he would do after the sale.
- 11 % → 12.8 %EBITDA, a year after the start
- 4.7 % → 2.9 %overall breakdown rate of the plant
- 8 → 5.2weeks to onboard a new operator
The situation
Stéphane is 51. He runs a continuing training company he founded in 2008. In the summer of 2024, the business was running at a pace he had not seen for three years. The post-Covid period had boosted demand for management training. His leadership team, now nine people, had been asking him since February for a clear AI roadmap. He did not have one. On 14 July, on holiday in Vendée, his wife pointed out that he had not looked up from his phone through the whole dinner with their two children. A few weeks later, he heard me speak on decision-making under pressure, at an event organised by a business club. He wrote the next day.
The Leader’s Brain assessment took place over one day in late November 2024. Two things came out of it: an attention system permanently saturated, and a decision pattern that, 80 % of the time, favoured avoiding conflict with his long-standing managers. In practice, Stéphane had been putting off the appointment of an operations director for fourteen months because he did not know how to announce it to two of the old guard. The three-day AI assessment inside the company, in December, identified four workable use cases. The most mature technically: automatic qualification of inbound leads, which his sales team was handling by hand sixteen hours a week. Estimated return on investment: 80 k€ a year, plus a qualitative effect on conversion.
What we did
- Fourteen individual sessions, one every three weeks, quarterly framing notes, and the direct line, which Stéphane used seventeen times, often at the end of the day.
- On the AI side, the company had no IT director: I selected an independent technical partner and coordinated the deployment of the first use case in four months, using n8n connected to ChatGPT Enterprise.
- A second use case, automatic drafting of commercial proposals from templates, went into production in month eight.
The hard moment
In March 2025, after a week in which he had nearly lost a 400 k€ client, Stéphane announced that he wanted to stop the engagement: "the coaching was doing nothing". I called him back on a Saturday morning. The conversation lasted forty minutes. He stayed.
"The coaching was doing nothing."
Six months on
In May 2025, the first use case had been running for six weeks. The sales team had recovered 13.8 hours a week, fewer than the sixteen targeted, but enough to move two people onto closing sales instead of qualifying leads. Stéphane’s watch showed an average of 6 h 40 of sleep, against 5 h 20 at the start.
A year on
A year after the start, the figures had consolidated. The first use case generated 73 k€ of annualised savings, slightly below the initial estimate, but with an unplanned side effect: conversion on inbound leads rose by 6 %, because the team called back highly scored prospects within the hour. The second use case freed 4.2 FTE (full-time equivalents) in sales from writing quotes; gross margin on the key-account segment gained three points. And then, in July 2025, Stéphane appointed his operations director. He also got into the habit of leaving at 6 pm on Fridays. His eldest daughter sat her baccalaureate in June; he was there.
- 5 h 20 → 6 h 40of sleep on average, six months on
- +3 pointsof gross margin on key accounts
- July 2025appointment of the operations director, long postponed
The situation
Anne-Sophie co-founded her company ten years ago with Julien, her chief technology officer (CTO). They raised 18 M€ in November 2023 from two European funds. The company was growing, ARR was rising, the team had doubled in eighteen months. And since the first quarter of 2024, nothing had been working between them any more. The board, with the investors, was pushing for a fast pivot to the healthcare vertical. Anne-Sophie felt that the technical debt of the product would not allow them to keep that commitment. Julien argued the opposite. Leadership team meetings froze. Two directors resigned at the end of 2024, both saying the same thing in private: "impossible to know who decides". ARR growth, running at +12 % in the fourth quarter of 2023, had fallen to -3 % in the fourth quarter of 2024. Anne-Sophie contacted me in January 2025, on the recommendation of a professional coach she had been seeing for three years, who had told her: "I think you need someone who also speaks the language of tech".
The Leader’s Brain assessment, in February, gave a diagnosis nobody had named: her decision system had switched to binary mode (Julien is right / I am right), with a saturation of the prefrontal cortex showing up as broken nights and solitary work in the evening, from 10 pm to 1 am, to compensate. The three-day AI assessment brought up a fact that neither Anne-Sophie nor Julien had identified: 22 % of the product decisions taken in the fourth quarter of 2024 had been taken without Julien, on the basis of usage reports generated internally. I got them both to agree to look at that figure together. It was the first time they talked on the basis of the same data.
What we did
The work combined individual sessions and three-way sessions.
- Anne-Sophie had twelve sessions over nine months, and I ran three three-way sessions with Anne-Sophie and Julien.
- On the AI side, two use cases were deployed: a document assistant (RAG) over the product base and the sales documentation, which allowed the two founders and their leadership team to reason on shared facts; and automation of the monthly reporting to the board, which took Anne-Sophie nine hours every month.
The hard moment
In June, during a three-way session, Julien offered to resign. Anne-Sophie refused. In July, they signed a written six-page arbitration pact defining each one’s scope of decision.
"He and I are still holding together. That was what was at stake."
Six months on
In August 2025, no further resignation had been recorded in the leadership team since the start. The document assistant was used by 23 of the 32 people in management. Two board meetings out of three had gone by without visible tension: Anne-Sophie was keeping count herself.
A year on
A year after the start, the healthcare pivot was launched in October 2025 with an MVP (minimum viable product) that Julien considered acceptable and Anne-Sophie considered defensible. ARR for the fourth quarter of 2025 showed +9 %, against -3 % a year earlier. Gross margin held. But what Anne-Sophie wrote to me in November, in a short message, was something else: "he and I are still holding together. That was what was at stake". She no longer works on Sundays. She has taken up tennis again, which she had stopped in 2022.
- -3 % → +9 %ARR growth, from one fourth quarter to the next
- 0further resignations in the leadership team, six months on
- 6 pagesof arbitration pact signed by the two founders
The situation
In 2008, the new-build housing market collapses. Developers sell around 40 % fewer homes: 127,400 in 2007, between 75,000 and 78,000 in 2008. Banks will no longer finance a development until around 70 % of the units have been sold off-plan.
The company builds the traditional way, in concrete. Its costs and lead times are those of the whole market. In a crisis like this one, doing what everyone does means sinking with everyone.
What we did
- A new product. We found a building system based on refrigerated panels, a clean break from concrete.
- An industrialised process, which halved the production cost.
- Land: a systematic search for cheaper plots.
- Local partnerships to develop new residential schemes.
The result
For 35 single-storey homes, we were weathertight and airtight in two months. The development could be delivered in six months in total. While the crisis was at its height, between 2008 and 2009, the company’s revenue grew by 30 %.
- +30 %in revenue, in the middle of the crisis
- -50 %in production cost
- 2 monthsto make 35 homes weathertight and airtight
What I took from it
A crisis does not reward the most cautious. It rewards those who change their product, their methods and their partners while the others wait. I lived it as a CEO. That is the perspective I bring to the leaders I work with.
Sources for the context: Fédération des promoteurs constructeurs, quoted by Batiweb (2007 and 2008 sales); French Senate, opinion on the 2009 finance bill (off-plan pre-sale threshold).
Shall we talk?
Thirty minutes on the phone. You set out the situation, I tell you frankly whether I can be of use.