Understanding the real need behind the request. A client who asks for "a report" rarely wants an Excel file — they want to know if their campaign is working.
- AXA 2022-2024: Media teams could only assess performance at the end of a campaign. I moved reporting to a weekly cadence — they could steer in real time, not after the fact.
- FEI: The federation wanted to monetize its sites. I recommended not activating the Ad Exchange to protect their premium positioning. Four years later, that decision still holds.
- My Media: Every advertiser had different KPIs. For AdopteUnMec, it was profile segmentation. For Sofinco, validated leads. I tailored the tracking to each business model.
- Le 10 Sport: Designing an ad architecture that generates €10-30k/month while preserving user experience.
When I take on a subject, I take it end to end. I don't stop at the edges of the job description.
- AXA 2022-2024: Two years on the Adverity project, from initial setup to team autonomy. The goal wasn't to stay indispensable — it was for them to no longer need me.
- My Media: Built the tracking department from nothing — recruitment, processes, tools, P&L. Nobody asked me to manage the P&L, but if you don't, you don't know whether the unit is profitable.
- IPG Head D&T: In 2017, I was an executor. In 2020, they called me back to manage the team. If they call you back after 3 years, the first time went well.
- Le 10 Sport: 5 years of full responsibility for monetization. Complete autonomy.
Before proposing a solution, I understand the problem. That means getting into the data, verifying assumptions, asking the right questions.
- AXA 2022-2024: The Adverity matching rules weren't working on certain edge cases. I had to dig into the naming conventions, the exceptions, the cases where the format didn't fit — before finding the right logic.
- MEC Blizzard: Auditing 3,700 tags to understand why certain campaigns weren't tracking correctly. The issue wasn't the tags — it was the absence of a coherent naming convention.
- AXA 2016: Auditing Weborama account structures country by country. Each country had its own logic. Everything had to be mapped before I could propose a unified convention.
- My Media: For Sofinco, we didn't measure leads — we measured validated leads. That required understanding the client's validation process end to end.
The quality of the plan doesn't count if the result isn't delivered. I measure my work by what's in production, not what's on the PowerPoint.
- AXA 2022-2024: Campaign go-live timelines reduced by 50%. 100% of Media Plans standardized. Daily reprocessing operational.
- MEC Blizzard: Successful launches across 27 countries for every major extension. Zero incidents despite the complexity.
- AdNovia/PHD: +300% campaign volume in 3 months. When I arrived, they had one client. When I left, that client was entrusting them three times more.
- My Media: 20% margin improvement over 2 years.
If a process is repetitive, automate it. If a tool doesn't exist, build it. True sophistication is when it's simple to use.
- AXA 2022-2024: A single Media Plan template replaced dozens of different formats across 30+ entities. Simple, but it changed everything.
- MEC Blizzard: I built a tag builder and monitoring system to manage 3,700 tags. Before that, every tag was checked manually.
- My Media: No API available? I wrote an email parser to automatically pull reporting data. Not elegant, but it worked.
- Starcom: Automated reporting tool using Weborama and DFA APIs.
A misconfigured tag means a flawed business decision. Data quality isn't a luxury — it's a prerequisite.
- AXA 2022-2024: Strict naming conventions and systematic quality controls. No campaign goes live without verification.
- AXA 2016: Insisted on creative and tracking integration compliance before every launch.
- My Media: Site-centric vs ad-centric measurement discrepancies were treated as inevitable. I designed a tagging architecture to reduce them.
- Mindshare: When I took over the MAIF account, I spotted errors in the existing reports. Fixed within the first week.
Trust isn't declared. It's built delivery after delivery, over time. When a client calls you back 3 years later, that's trust.
- Le 10 Sport: 5 years working with the CEO. That doesn't last unless the work is reliable.
- FEI: 4+ years of continuous service. Same contact, same standards, same trust.
- AXA 2022-2024: Built trust with 30+ internal entities — as an external consultant with indirect management authority.
- IPG Head D&T: Called back in 2020 to take over team management. The trust established in 2017 had held.
Training someone isn't showing them buttons. It's understanding their level, adapting the pace, and making sure they're autonomous when you leave.
- AXA 2022-2024: The Campaign Management GBS team (3-4 people) went from dependent to autonomous over 2 years. The real test: can they run without me? Yes.
- AXA 2016: Training the offshore India team — remotely, in English, across time zones. That requires clear documentation and systematic feedback.
- My Media: Recruited and trained an ad ops manager. When I left, the department ran on its own.
- IPG Head D&T: Team of 4, mix of junior and experienced. The job was as much about shielding the team from unstructured requests as it was about helping them grow.
An imperfect pilot beats a perfect plan that never launches. Start, measure, adjust.
- AXA 2022-2024: Rolled out the Media Plan template country by country, without waiting for perfection. Field feedback improved the template along the way.
- Le 10 Sport: No ad infrastructure existed. I configured GAM and AdExchange quickly to start generating revenue.
- Ad Ops Missions: At every new assignment, operational immediately. No observation period when a client is waiting for their campaigns.
Ad tech moves fast. If you don't keep up, you're obsolete in 2 years. Every engagement taught me something.
- AXA 2022-2024: I organized Server-to-Server, CAPI and Privacy Sandbox workshops. Not to make slides — to concretely test solutions before third-party cookies disappear.
- MEC Blizzard: In 2014, DMPs were emerging. I learned RocketFuel and X+1 on the job, while supporting Blizzard's exploration.
- Le 10 Sport: Proactive testing of Sublime Skin, Taboola, Criteo, Adikteev — to diversify revenue sources, not out of idle curiosity.
Solve today's problem, but design the solution so it holds tomorrow. Think scale from the start.
- AXA 2022-2024: The econometrics pilot for France wasn't just a one-off project — it was a model designed to be deployed in other countries.
- FEI: The monetization strategy designed in 2021 still holds in 2025. That's because it was built to last, not to maximize the short term.
- MEC Blizzard: The DMP integration was a vision for data transformation at a global gaming company.
Optimize before you buy. The best solutions aren't always the most expensive.
- AXA 2022-2024: The Adverity migration reduced maintenance costs vs the previous system. Centralized license management generated group-wide savings.
- Le 10 Sport: Revenue maximization through passback chaining and floor price testing. Didn't need more inventory — just better use of what was already there.
- MEC Blizzard: Tag builder and monitoring developed with Excel and existing Sizmek data. Zero additional tool budget.
Recommend based on data, not gut feeling. And when the data shows you were wrong, own it.
- AXA 2022-2024: The Ad Verification RFP was conducted with objective scoring criteria. IAS vs DoubleVerify benchmarking — recommendation based on tests, not preferences.
- Le 10 Sport: Every recommendation balanced business constraints (revenue), technical constraints (SEO) and UX. No perfect solution, but the best available combination.
- FEI: The decision between direct sales and Ad Exchange was based on analysis, not dogma. Four years later, the numbers validate it.
Say what you think, even when it's uncomfortable. Then respect the final decision and commit fully.
- Voyages-sncf: I recommended Ad Exchange integration. Management declined — institutional constraints. I learned that day that the best technical solution isn't always the right business decision. And I moved on.
- Le 10 Sport: Recurring tension between CEO (revenue) and tech team (SEO). I defended my recommendations, but when SEO took priority, I aligned.
- AXA 2022-2024: Defended my recommendations during the Ad Verification RFP, even when they went against initial preferences.
