Challenge
Predicting customer behaviour can give a company the edge in a fierce and competitive market.
That’s why this luxury automotive company company asked us to design a model that could predict when its customers were likely to buy a new car. They had two goals for the project: they wanted to see an increase in revenue and customer loyalty.
The number of cars sold in the company’s market is relatively low. Even a small degree of additional sales matters and can be the difference between meeting and missing sales targets.
Profusion were asked to design a model that could predict when its customers are likely to buy a new car – and so increase revenue and customer loyalty.
Solution
We aimed to expand and unify the company’s data points to gain better insight into their customer base. So, we:
Designed a de-duplicating algorithm that cleansed the customer base and harmonised its data, allowing for sharper and more targeted interrogation
- Designed a de-duplicating algorithm that cleansed the customer base and harmonised its data, allowing for sharper and more targeted interrogation
- Found a way to collect higher quality data — greater data accuracy leads to more accurate models
- Created a tool to capture additional data points and provide an expanded basis for analysis.
- Used our data model to increase conversations by recommendation within Salesforce.
Impact
Our bespoke de-duping algorithm spotted over 5,000 duplicates with above 99% accuracy, leaving the company with a more robust and rigorous dataset.
Using this expanded and enriched data, our customer targeting engine could predict the exact time period in which the next purchase is made with 75% greater accuracy than any previous attempt by our client. After just three months, our recommendation system has generated more than a fifth of the company’s opportunities and contributed over £13 million in sales.
Our partnership has led to an increase in sales opportunities and conversions for the company, giving them the edge in their high-stakes market.