The digital age has changed the way the modern consumer shops – not only do we continue to see an increase in e-commerce but, more and more, people are using online research to inform what they want to buy in the first place. Understanding the impact of marketing on overall sales has been a long-standing challenge but this pain is acutely felt in the automotive sector where, although research may have been carried out online, the purchasing of a car is still done in person. Linking your online marketing activity to your sales become near on impossible.

This is what our client, a major UK Media Agency, wanted to understand better. As advisors to a number of automotive companies, their hypothesis was that a the most influential stage of a car buyer’s decision making process was during online research. Our client wanted not only to see if they could prove this but also to understand the online behaviours that can lead to a purchase.

What we did

Coppelia used a technique called Agent-Based Modelling; a way of simulating the decision making process of buying a car. This model allowed us to model the factors that would influence a) whether they would buy a car and b) what type of car they would buy.

Using research and historical sales data, Coppelia was able to understand which stages of the buying a car were the most important and, therefore, where marketing has the greatest opportunity to make an impact.

The outputs of this model were presented using glasseye, Coppelia’s own interactive visualisation capability. This allowed board members not only see the results but run through different scenarios to decide the best way to adjust their marketing.


This tool gave our client a tangible means to challenge current views about buying decisions and also show the potential gain of changing their marketing strategies. Overall, our client, as a media agency, were better equipped to advise their automotive clients on the optimal mix of media to achieve their sales objectives.

Why Coppelia?

Coppelia can not only build useful and credible market simulations, we can also help you understand their true value. Simulations are an extremely powerful way to communicate and investigate ideas since they deal with the concrete things in our world rather than mathematical abstractions. We can help you harness their explanatory power to bring about change in your organisation.

What is Agent Based Modelling?

Agent-based modelling involves simulating the behaviour of people (agents) as individuals in complex environments. In our case these simulated agents are programmed to have the qualities and behaviours that online shoppers have and the simulated environment consists of the automotive markets and the online purchase funnel. Some of the behaviours are informed by research others are are hypotheses that we have and we will test. Once we have set up our simulated world we can play out various scenarios to work out what would happen, for example we can change a marketing strategy.

Machine Learning and Analytics based in London, UK