AI & Data Science Expert · Market Research

Turning data into evidence for market research.

Jean-Noël Zeh, Chief Data Officer, 25 years of experience in scientific direction. Specialist in data fusion, machine learning and agent-based modeling applied to market research and audience measurement.

Portrait of Jean-Noël Zeh
25 years
of scientific direction
4M+
synthetic individuals modeled
1.5Bn+
synthetic trips / year
600,000+
billboards valued
15,000+
GPS panelists tracked
Background

From software engineering to scientific direction

A career built on the combination of technical and methodological expertise, at the service of market research.

  • 1986
    Engineering degree
    École Centrale de Marseille.
  • Early career
    Software manager
    Design of data processing and analytics solutions.
  • Before 2018
    Scientific Director — CSA Institute
    Scientific direction of studies, methodology and data innovation.
  • Since 2018
    Chief Data Officer — Ipsos
    Leading the Data Management Center (DMC) and its international deployment.

Strong expertise in data science and applied statistics: data fusion, machine learning (driver analysis, segmentation with digital activation), mobility and behavioral modeling.

At Ipsos, Jean-Noël Zeh leads the Data Management Center, the entity in charge of designing and deploying data and AI solutions for market research, audience measurement and customer insight.

He notably leads Mobimétrie, the reference audience measurement currency for out-of-home (OOH) advertising in France, presented at the ESOMAR 2024 world congress (Athens) and recognized with the Grand Prix Syntec des Études (2023 and 2024 editions) — a project that has mobilized close to 8 person-years over more than 4 years, with methodologies now deployed outside France (United Kingdom, Norway) and under evaluation in Japan, Ireland, Switzerland and Germany.

His expertise spans the full data value chain: from collection (panels, passive data, open data) to advanced modeling (Agent-Based Modeling, synthetic data), through to large-scale cloud computing (Google BigQuery, VertexAI).

Expertise

State-of-the-art methods for the service of insight

Expertise built on real, national-scale projects, combining scientific rigor with operational demands.

01

Data Fusion & Machine Learning

Driver analysis, segmentation with digital activation, predictive models applied to behavior and consumption.

02

Agent-Based Modeling

Simulation of synthetic populations and individual behaviors to reconstruct collective dynamics at scale.

03

Synthetic Data & Virtual Populations

Generation of representative populations calibrated on public data (INSEE) to overcome the limits of self-reported data.

04

OOH & DOOH Audience Measurement

Modeling of traffic flows and advertising visibility, outdoors and indoors (retail, transit, offices).

05

Large-Scale Cloud Computing

Industrialization of data pipelines on Google BigQuery and VertexAI to process billions of data points.

06

Generative AI Applied to Research

Integration of generative AI into analysis, methodological writing and market research reporting workflows.

07

AI-Calibrated Market Sizing

Market sizing models combining machine learning and generative AI to strengthen and refresh market size estimates.

08

Research Production Automation

Design of report and presentation automation tools, integrating generative AI to accelerate research reporting.

09

Training & Knowledge Transfer

Design and delivery of internal training in applied statistics, data integration and segmentation, upskilling research and data teams.

Work

Projects that set the standard

A selection of landmark projects, recognized by the research and advertising industries.

Flagship project

Mobimétrie — France's reference OOH audience measurement currency

Design and scientific direction of the model that estimates the value of each of France's 600,000+ billboards, by simulating the movements of more than 4 million synthetic individuals.

  • Methodology based on Agent-Based Modeling: reconstructing trips individual by individual (home, work, leisure) rather than an aggregated flow model.
  • Calibrated on more than 15,000 GPS-tracked panelists, drawing on 100+ data sources organized into 6 families (surveys, passive GPS measurement, transport data, INSEE open data, pedestrian counts, telecom data).
  • Unprecedented granularity: detailed socio-demographic profile across 20 variables, 50+ trip purposes modeled over more than 10 million street segments.
  • Industrialized on Google BigQuery and VertexAI to generate more than 1.5 billion synthetic trips per year.
  • Presented at the ESOMAR 2024 world congress (Athens); winner of the Grand Prix Syntec des Études, 2023 and 2024 editions.
600,000+
billboards valued
4M+
synthetic individuals simulated
1.5Bn+
synthetic trips generated / year
87%
weekly coverage of agglomerations with 10,000+ inhabitants (source: IREP, 2026)
2026-2027 rollout

Indoor Audience Measurement — a first in France

Extension of the Mobimétrie methodology to indoor environments (retail, gyms, campuses, office buildings, then transit and airports), with an individual-level granularity unprecedented on the French market.

  • Two Proofs of Concept already validated: with Doohyoulike across 2,005 venues (gyms, campuses, local retail) and with ECN across 210 office buildings and 308 screens, with a target-to-result gap below 1%.
  • Phased 2026 rollout: first "simple indoor" environments before summer, more complex environments (shopping centers) next, transit and airports by early 2027.
  • Same reference synthetic population as the outdoor model, ensuring cross-environment comparability.
  • Nomination submitted to the Adwanted Research Awards (IREP) 2026, in partnership with Ipsos bva, for this first individual indoor audience measurement in France.
< 1%
profile accuracy gap
2
Proofs of Concept validated (Doohyoulike, ECN)
2027
transit & airports horizon
European project

EONA-X / System-X — Synthetic mobility for carbon-neutral cities

Scientific contribution to a European project (under the IPCEI AI framework) mobilizing synthetic mobility data to support the climate transition of urban and peri-urban territories.

  • Hierarchical modeling combining GPS panels, transit data (GTFS), pedestrian counts and telecom mobility data.
  • Simulation of "what-if" scenarios: car-free city centers, impact of electrification, modal shift.
  • Multi-year longitudinal tracking to measure the real impact of public policy on mobility and climate objectives.
5Bn
synthetic trips modeled
6M
street segments analyzed
3 years
of longitudinal analysis depth
Talks & Awards

Speaking up, and being recognized by the industry

Speaking engagements at professional and scientific conferences in France and internationally, and recognition of the Mobimétrie project by the research and advertising industries.

Conferences & seminars

  • Sept. 2024
    ESOMAR Congress — Athens (Greece)
    International talk "Mapping Billions of Synthetic Journeys", followed by an Ipsos debrief webinar in October 2024.
  • 2024
    Colloque des Sondages
    Paper "Fine-grained modeling of out-of-home audiences", on data fusion and multi-agent modeling.
  • 2024
    CMI Community — Ipsos
    Talk on generative AI and synthetic data applied to market research.
  • June 2024
    EMRO
    "Innovating in OOH measurement", on the Mobimétrie audience measurement methodology.
  • Oct. 2023
    Ipsos client seminar
    Presentation of the Mobimétrie methodology to advertisers, media agencies and outdoor advertising companies.
  • May 2025
    Segmentation & AI Webinar — Ipsos DMC
    Talk on segmentation use cases activated through generative AI.
  • Oct. 2025
    IREP After — Synthetic Data
    Presentation of synthetic data use cases for market research.
  • Nov. 2025
    IRT SystemX / EONA-X
    Presentation of the European ATLAS project on synthetic mobility for carbon-neutral cities.
  • June 30, 2026
    IREP Forum · UDECAM
    Presentation of the unified Mobimétrie × Ipsos bva OOH Indoor model, a granular, individual-level audience measurement.

Awards

Winner

Grand Prix Syntec des Études

2023 and 2024 editions, for the Mobimétrie project: building a synthetic trip database using Agent-Based Modeling (ABM).

Selected

ESOMAR Congress 2024

Selected for an international talk in Athens at the 2024 edition of the ESOMAR world congress, the leading conference of the market research industry.

2026 nomination

Adwanted Research Awards — IREP

Nomination submitted in 2026 for the indoor extension of Mobimétrie, in partnership with Ipsos bva; a co-authored op-ed was published in The Media Leader. Result not yet known.

Press

In the news

A selection of trade press coverage of Mobimétrie and its recent developments.

Methodological Foundations

Methods grounded in the scientific literature

The approaches used — agent-based modeling and synthetic populations — draw on an established academic corpus in data science and transportation research.

Agent-Based Modeling Bonabeau, E. (2002). "Agent-based modeling: Methods and techniques for simulating human systems." Proceedings of the National Academy of Sciences, 99(suppl 3), 7280–7287.
Synthetic populations Beckman, R. J., Baggerly, K. A., & McKay, M. D. (1996). "Creating synthetic baseline populations." Transportation Research Part A: Policy and Practice, 30(6), 415–429.
Population synthesis for microsimulation Müller, K., & Axhausen, K. W. (2011). "Population synthesis for microsimulation: State of the art." Transportation Research Board 90th Annual Meeting.
Data fusion in market research Baker, K., Harris, P., & O'Brien, J. (1989). "Data fusion: An appraisal and experimental evaluation." Journal of the Market Research Society, 31(2), 153–212.
Services

How we can work together

Tailored support, from methodological design to industrialization, for research institutes, advertisers and advertising networks.

01 Methodological consulting

Design of data fusion methodologies, agent-based modeling, synthetic population generation and AI-calibrated market sizing tailored to your research challenges.

02 Scientific direction of data projects

Scientific direction of audience measurement, segmentation or predictive modeling projects, from specification to results validation.

03 Talks & conferences

Speaking engagements at professional conferences (ESOMAR, industry symposia) on AI, synthetic data and the future of market research.

04 Training & workshops

Upskilling research and data teams on machine learning, applied statistics, data integration, segmentation and generative AI.

Let's discuss your project

A methodological question, an audience measurement project, or a speaking engagement to consider? Let's get in touch.