Case Study · Predictive Analytics

Recommendation System for Automotive Products & Services

A hybrid recommendation engine that suggests the next best service, accessory, protection product or vehicle upgrade for each customer, shown to service advisors, in the customer app and in email.

← All case studies

Client

Automotive retailer with service, accessories and F&I products

Industry

Automotive

Delivers

Personalised next-best offers at every touchpoint

The problem

Service advisors had seconds to suggest extra work, accessories or protection plans, and relied on memory and generic promotions. Online, every customer saw the same offers whatever they drove or when they had last visited.

Approach

A hybrid model: collaborative filtering on purchase history captures what similar owners buy, and content-based signals cover vehicle model, age, mileage, season and service schedule.
Business rules guarantee compatibility (parts fit the vehicle), respect inventory and margin, and suppress items bought recently.
Offline evaluation with precision@k, recall@k and catalogue coverage, followed by online A/B tests before full rollout.

Where recommendations appear

The service advisor screen when a repair order is opened.
The customer app and online service booking.
Pre-visit and post-visit emails.
Upgrade suggestions for sales teams when a customer's vehicle reaches its likely replacement point.