Five real projects I led.

Context

When I took over the segment, I was the only AE on the French public portfolio. UGAP had just expanded its language framework, the project portfolio was growing, but we had no shared methodology, no standard tooling, no profile to hire against.

Problem

Hire three Account Executives in 12 months capable of running long cycles (6-18 months), multi-stakeholder, in a complex regulatory environment. The right profiles for public sector look nothing like private-sector hires: fewer hunters, more patient farmers.

Approach

• Defined a hiring grid with 3 filtering questions (published in this article)
• 90-day onboarding plan in 3 phases (immersion / accompanying / autonomy)
• Internal training on centrales d'achat mechanics with documented playbook
• Non-negotiable weekly 1-1s and post-call client debriefs using the GROW model

Outcome

Three AEs hired and onboarded over 12 months, no attrition. Operational productivity reached at 4 months on average (vs 6+ in similar roles elsewhere). Shared sales methodology across the team, qualifiable in weekly review through MEDDPICC adapted to public sector.

Context

The French public sector is enormous — central administrations, ministries, operators, local authorities, hospitals, universities, agencies. Without prioritization, the team spreads thin and converts little.

Problem

How to score 1,200+ accounts so an AE knows by Monday morning what to prospect, with which angle, and what value to expect?

Approach

• Excel MasterFile structured by typology (central, operator, local government, healthcare, education)
• centrales d'achat eligibility score based on recent public procurement
• Priority score (size, training budget, weak signals like RAP/Bilan Social)
• Pre-written, adaptable commercial angles
• Scaled tests on pilot accounts (MEAE, GRDF, SNCF, Ministère de la Culture, Institut Pasteur)

Outcome

Tool used daily by the AE/AM team. Qualification time cut by ~60%. First qualified contacts on 14 test institutions, several mature deals at 6 months.

Context

2025 marked the AI tipping point in public sector: France's DINUM is pushing Microsoft Copilot and sovereign AI platforms La Suite Numérique, ministries are deploying Copilot, local governments are questioning their stack. Language training must fit into this ecosystem.

Problem

Build a technically credible pitch and offering for CIOs and CHROs who no longer accept generic AI talk. How to position goFLUENT as a strategic AI partner, not a course vendor?

Approach

• Mapping of integration possibilities (Microsoft Copilot and sovereign AI platforms, La Suite Numérique) with their training use cases
• Built a 15-slide deck integrating Cegos Barometer 2026 data and our AI vision
• Prepared sovereignty arguments for ministerial accounts (French hosting, GDPR, data)
• Piloted with two test institutions to iterate the pitch

Outcome

Pitch validated internally and used on strategic 2026 meetings. Two deals in instruction on the ministerial segment. Progressive positioning of goFLUENT as an AI-native player in public training.

Context

The team defaulted to BANT — fine for classic B2B, ill-fitted to public sector where cycles involve 5-7 decision-makers (prescriber, public buyer, decider, finance, end user, legal).

Problem

Equip the team with a shared methodological frame fitted to public-sector reality, without losing 6 months to theoretical training.

Approach

• 12-section Word document, in French, with concrete examples from real centrales d'achat cases
• Comparison BANT / MEDDPICC / SPIN / Challenger Sale with usage recommendation per cycle phase
• Repositioning goFLUENT as strategic HR partner rather than course vendor
• Discovery scripts specific to public buyers (open questions adapted)

Outcome

Playbook adopted by the team and used in weekly project portfolio reviews. Improved qualification — fewer ghost deals in project portfolio. First version published internally, already two iterations.

Context

Competitive intelligence took my team two days a month — manual SIRENE research, competitor site scraping, manual synthesis. Not scalable.

Problem

Automate intelligence to retrieve financial and legal data per competitor, and generate exploitable synthesis in 30 seconds.

Approach

• Vanilla HTML/JS web app, hosted internally
• SIRENE connector for financial and legal data
• Claude API to summarize data into readable analysis (positioning, strength/weakness)
• Dashboard with multi-criteria scoring (Price, Technical Value, Service Quality, PMDD)

Outcome

Intelligence time cut by 10x. Battle cards auto-generated and used in team meetings before any strategic deal. First concrete AI application in my team — which also fed my AI credibility in client pitches.