How Pierre Fabre trained 60 reps with AI. And why sales training will never go back to what it was.
In 2022, Pierre Fabre Medical Care France made a bet no French laboratory had made before: reshape the sales culture of its 82 medical reps from the ground up, then plug in an AI to train every rep on their real selling behaviors.
Four years on, the partnership is still running. Three assignments back to back. And one line from the Pierre Fabre project lead that says it all:
"The other vendors offered standard, cookie-cutter stuff. b-flower brought Pierre Fabre into the thinking, while the other firms imposed their own models."
Here is what happened, step by step. And why it prefigures what is about to become the norm in B2B sales training.
Step 1The realization nobody wanted to make
In 2022, French pharmacy is going through a shift. Foot traffic is up 9.6%. The pharmacist becomes a central player in the care pathway. And the medical rep's job shifts with it: you no longer sell volume, you manage shelf space, you train pharmacy teams, you run a merchandising diagnosis.
Except the sales teams were still trained the way they had been twenty years earlier. An annual seminar, role-plays debriefed by gut feel, a sales playbook updated every 18 months. Nobody thought it was enough. Nobody had an alternative.
Pierre Fabre refused the easy answer ("let's hire more senior profiles"). They accepted that the problem was not about people, it was systemic. That mental shift is what makes everything else possible.
Step 2Co-build the method, don't impose it
b-flower is brought on as a strategic partner. And here comes the first real difference with what 90% of sales-training firms do.
Instead of showing up with a standard methodology to roll out, b-flower spent the first months listening. Ride-alongs with 18 reps in the field. Interviews with pharmacist customers. Workshops with the three internal departments (marketing, sales, omnichannel). The sales method wasn't handed over. It was built with Pierre Fabre, from their field, their portfolio, their personas.
A co-built method takes root. An imported method rolls out for six months, then dies. Every sales director who has lived through a "big transformation project" that ended up in a drawer knows exactly what we mean.
Step 3Train differently depending on the resistance
Pierre Fabre knew some of the reps would push back. Rather than pretend everyone would get on board at the same pace, b-flower segmented the training tracks by level of resistance to change.
The enthusiasts moved fast. The skeptics got awareness workshops before the technical tracks. The managers were trained a step ahead, so they could carry the message and coach in the field from day one.
Most transformation projects fail because they treat everyone the same. By segmenting on resistance, Pierre Fabre avoided the "false start" effect that kills 70% of sales-change programs.
Step 4Equip the field with the right tool
In parallel, Pierre Fabre was rolling out a new iPad CRM for its reps. That CRM plays exactly the role SalesApps plays today across the group's clients: put the right content in front of reps at the right moment. Without support, a tool like that gets rejected, underused, or worse, used the wrong way. b-flower built the rollout straight into the training tracks. The tool didn't land as a mandate from headquarters, but as the natural extension of the new sales posture.
Today, SalesApps puts the right content in every rep's pocket at the right moment: up-to-date product sheets, selling points by persona, prepared objections, sector-specific case studies, presentation decks, demo videos. The rep reaches all of it in the meeting, without leaving the tool, without digging through three SharePoint folders. Marketing and sales leadership update the content continuously, and every rep has the latest version of the message, in real time.
A rep who walks into a meeting with the current product sheet, the three prepared objections and a case study from the same sector is a rep who carries the message marketing built, instead of improvising a rough version of it. Tool plus method is what makes the transformation hold over time.
Step 5AI to close the loop
Two years into the project, the method was live, the reps were trained, the managers were coaching. But one very specific risk appeared: post-launch fatigue. Three stubborn blind spots:
- Training debriefs still rested on the trainers' subjective observation. Two trainers debriefing the same role-play didn't reach the same conclusions.
- Managers saw their teams' results, not the behaviors producing them. They knew a rep was slipping, without knowing exactly what to help with.
- Nobody measured what actually happened in customer meetings. The CRM captured whatever the rep chose to put in it, not what really played out in the room.
That's where Marvin entered the picture. Concretely, here is what we do at Pierre Fabre with the tool.
Each rep runs sales conversations in a trio: a rep, someone playing the customer, and Marvin observing. The conversation is recorded on video. Marvin then produces a full conversation analysis covering the relational micro-skills: talk/listen ratio (top performers talk 43% of the time on average, the market average is 68%), quality of discovery questions, reframing the customer's need, handling silences, objection handling, buying signals caught or missed, closing on an open question versus a pitch.
The b-flower coach picks up that analysis and debriefs each rep on a factual basis instead of an impression. The rep leaves with a detailed individual report. Their manager gets the same report. And every week, Marvin produces a map of the whole team: who is progressing, on which behaviors, at what pace. Not a results dashboard. A map of the behaviors that produce those results.
Today the setup goes further still: Marvin can analyze real customer meetings, not just simulations. The rep turns the tool on before a meeting, Marvin listens, produces an individual report the next day, and the manager sees for the first time what really happened in the meeting. Not what the rep chose to log in the CRM.
For the first time, sales leadership has an objective map of its network's relational skills. Not a warm satisfaction score. On that data, a manager can coach precisely, at the right moment, on the right behavior. Training no longer stops when the seminar ends. It runs continuously, 52 weeks a year. And when it's time to forecast, that changes everything: the numbers no longer rest on the rep's optimistic take on how their deals are progressing, but on objective signals captured in real meetings. The forecast finally becomes reliable.
The full equation: why the three together are worth ten times each one alone
The Pierre Fabre case is the first full-scale demonstration of an equation that is about to become the norm in sales performance. And that equation rests on real synergy between three building blocks, where each one makes the other two better.
Builds the method with the client and trains the teams. This is the human intelligence, the field knowledge, the co-building. 25 years of expertise, 6,000 reps trained per year.
Equips reps with the right content, at the right moment, in the right format, built directly into their daily work.
Closes the loop: it trains continuously with AI, it analyzes the conversations, it maps every rep week after week, and it feeds the manager data they can act on.
The synergy works both ways. Every b-flower assignment feeds Marvin qualified data on the client's real sales method, which makes the AI more precise. Every Marvin analysis sends the b-flower consultants field signals that make their interventions more surgical. Every piece of SalesApps content used in a meeting is measurable by Marvin, which lets marketing finally know what actually works. And every Marvin insight can trigger the creation of new SalesApps content or a targeted b-flower intervention.
Taken separately, these are three solid tools. Together, they are a system where each block makes the other two perform better, week after week. Nobody knows how to copy that today.
What this changes for your sales force
If you're reading this and thinking "that's all well and good, but we have neither four years nor the Pierre Fabre budget," that is exactly why Marvin exists.
The training AI, the analysis of real meetings, the weekly skills map, the individual reports that give the manager the information they were missing, all of it is available today, without waiting four years for a cultural transformation.
Pierre Fabre is the proof that the equation works. Marvin is how it becomes repeatable.