Forward-looking simulation of automotive customer segmentations.

Anticipating tomorrow's buyer expectations

Designing a new vehicle model takes between 6 and 10 years. This raises a strategic question for car manufacturers: what will customer expectations look like by the time the vehicle finally reaches the market? We developed a forward-looking simulation tool to address this challenge.

The context

Since 2021, working from a large annual multi-manufacturer survey of new car buyers — we have built a customer segmentation based on attitudinal variables related to cars. Organised into 7 clusters at the European level, this segmentation is reassigned every year and helps to better understand customer expectations in order to imagine and design tomorrow's vehicles.

The challenge

How can we project the evolution of these clusters over a 6 to 10 year horizon? The idea: formulate hypotheses on the evolution of demographic, behavioural, attitudinal and automotive variables, then measure their impact on cluster composition.

Reference data

The project draws on the databases, covering the 5 major European markets (France, Germany, United Kingdom, Spain, Italy), with approximately 30,000 respondents per country per year. To ensure statistical robustness, we work on two consecutive years (2022 and 2023), which form our reference dataset.

Two families of variables

The tool relies on two carefully selected sets of variables:

The simulation variables — those on which evolution hypotheses are formulated: socio-demographic criteria (age, income, household composition, place of residence), automotive characteristics (size class, energy type, body type, price range), attitudinal variables (relationship to safety, technology, environment, autonomous driving, driving pleasure) and usage variables (commuting, weekend trips, holidays).

The descriptive variables — those that help to better understand the simulated clusters without entering the calculation itself: sociological profile, expectations, detailed usage patterns, brand loyalty, household equipment.

The outcome

An accessible simulation tool that allows users, country by country, to vary the distributions of the simulation variables and observe the impact on clusters. Results are delivered at two levels: per country, then aggregated across the 5 European markets.

Each simulation automatically produces a complete deliverable: a summary document recapping the hypotheses applied, a visual mapping of the groups and their evolution, the full set of results in graphical form, statistical tests comparing the simulation to the reference data, and a detailed data file.

Our contribution

This project illustrates a core capability of our work: turning a long-term strategic question — "what will my customers look like in 10 years?" — into an operational tool usable by marketing and product teams, built on reference survey data and a rigorous segmentation methodology.