BMS France: from 20% to 72% regular use of generative AI

Customer reference · Pharma · Headquarters in Rueil-Malmaison

BMS France: from 20% to 72% regular use of generative AI

The tools were already deployed, but usage was stagnating. In 22 months and six short waves led by ten internal Champions, the proportion of employees regularly using Copilot and the group's internal conversational assistant increased from 20% to 72% by the fourth wave (April 2025), with a stable satisfaction level between 4,7 and 4,8 out of 5 in each wave.

September 2024 to June 20266 waves99 workshops10 mushrooms (champions)Prompt Party and Agent Party

The starting point: access without use

By summer 2024, BMS France employees already had access to the generative AI tools provided by the group: Microsoft Copilot and an internal chatbot. The problem wasn't access, but usage. The assessment made in summer 2024 was clear: only one adoption initiative had taken place, the perception was very positive, but usage remained low due to a lack of experience, and generative AI had not yet been established as a collective priority.

The tools are deployed, the perception is positive, but actual usage remains low: without a support system, nothing will take off.

The objective is operational: to facilitate the adoption of the two tools to improve or accelerate daily tasks for approximately 450 headquarters employees across all departments, within an initial timeframe of six months. NextStart was selected based on a comprehensive proposal submitted in the middle of August and a supportive approach that empowers internal teams, rather than a top-down consulting model. The project will go to the executive committee in September 2024 for an initial three-month phase, comprising two rounds and ten Champions.

The device: a four-step mechanism, repeated six times

Each wave follows the same architecture, from the co-construction of use cases to the measurement of usage, which allows for comparison of the results from one wave to another.

01

Prepare

Two one-hour meetings with the project team to select six concrete use cases per wave, based on the needs reported by the business: writing, translation, indicators, summarizing verbatim, reports, creating PowerPoint presentations, analyzing files, then reasoning models, web research and deep research, Excel, image, audio and video.

02

Train

A half-day and then two-hour "starter pack" is used to train ten volunteer internal Champions from different departments, supported by an internal project team. The Champions are retrained before each wave.

03

Prompt Partythen Agent Party

45-minute workshops (one hour for theAgent Party), in person, in groups of about twelve, with voluntary registration from around twenty available time slots, always co-facilitated by a BMS Champion and a NextStart expert. In 2026, each participant leaves with a Microsoft 365 Copilot agent built during the session: a study analysis tool, a monitoring tool, or a "deliverable coach" that reviews and challenges a document before sending.

04

debrief

A feedback session with the Champions at the end of each wave, a satisfaction questionnaire, and usage measurements before and after. Around the workshops: a library of prompts and replays on SharePoint, Viva Engage facilitation, stickers, and regular internal communication.

The measured results

Regular use: 20% to 72%

Before the first Prompt Party20% of respondents use generative AI regularly, 60% occasionally, and 20% never. After waves 1 and 2 (December 2024): 47% regularly, 9% never. After wave 4 (April 2025): 72% regular use. Debriefings of waves 3, 4, and 5 reveal regular or occasional use at 90%, 90%, and 93%, respectively.

Satisfaction: 4,7 to 4,8 out of 5 in each wave

4,74 (128 respondents) and 4,81 (123) on waves 1 and 2; 4,71 (109), 4,79 (86) and 4,77 (81) on waves 3 to 5; 4,71 (104 respondents) on theAgent PartyFrom the first wave, 97% judged the use cases relevant and 100% rated the performance of the facilitators as "excellent or good".

Presence and demand

Attendance rate of 90 to 98% compared to registered participants. 220 registered for 200 places in wave 1, 110 registered in one day at the opening of wave 5, 134 registered in the middle of a school holiday week for the...Agent Party.

Agent Party One in five participants has already created an agent

At the end of wave 6, 20% of the 104 respondents reported having already created their own Copilot agent after the workshop. The program has been renewed three times by the Executive Committee, from the initial three-month wave to the 2025 program in three rounds, and then to the...Agent Party 2026.

Key figures

22 monthsprogram, from September 2024 to June 2026
6successive waves, 5 Prompt Party then 1 Agent Party
99In-person workshops: 84 Prompt Party and 15th Agent Party
10Internal champions trained and reformed before each wave
72%of regular use after wave 4, compared to 20% at the start
4,7 to 4,8/5satisfaction on each of the six waves

What they say

"We have succeeded in making PROMPT PARTY a real success, and it's time to celebrate this incredible adventure together."

Project team, BMS France, 2024

"A score of 4,79 is one of our best scores."

Project team, BMS France, 2025

What we take away from this

The short format wins

45 minutes, around twenty time slots, in person, six concrete cases per wave: these are the success factors listed by BMS from the very first debriefing. The time constraint forces practice.

The Champion + Expert duo

The Champion brings professional legitimacy and knowledge of internal tools, while the external expert provides methodology and insights into new developments. BMS cites this partnership as one of the keys to the system.

Measure at each wave

The measurement before and after each wave made it possible to show the Executive Committee a continuous progression and to obtain three successive renewals.

To advance usage, not just knowledge

Your tools are deployed but usage is stagnating? Let's talk about a short-wave approach, internal relays, and usage measurement tailored to your organization.

The same program for your organization

What BMS France has implemented is available in catalogue format: the AI ​​Ambassadors programDescribe your situation in a few lines, and we'll get back to you within 24 business hours with a proposal. You can also book a 30-minute consultation directly.


Do you prefer to call or write? +33 1 89 16 48 05 or hello@nextstart.aiYour contact details are only used to reply to you.

Explore more!

The formats used in this program and related references.

Back to top
The entire NextStart.AI libraryEight consulting missions, seven pillars, around twenty formats, fifteen professions, five tools
Nine detailed customer case studiesLarge groups, public agencies and networks: from acculturation to production

Insurance and social protection

Health and life sciences

Public sector and research

The NextStart.AI collectiveConsultants, trainers and industry experts: more than 4,850 professionals trained since 2023
All free resourcesDiagnostics, comparison tool, Finders, charter template, Word templates and industry-specific kits
The NextStart.AI blogComparisons of tools, practices, agents, governance: analyses and feedback from the field
Let's talk about your AI projectResponse within 24 business hours, written quote after initial discussion
AI agents in businessUnderstanding, choosing the tool, training the teams, deploying
Delegate entire tasksClaude Cowork, Copilot Cowork, ChatGPT Work, Vibe Work