Our client is one of the largest betting retailers in the UK, providing gambling and gaming entertainment products and services through retail stores and websites.
As a leader in the sector, their revenue depended on their active customers. To keep customers engaged for longer periods, the CRM team sent personalised marketing messages manually. However, as their customer base grew to millions of accounts, this process became inefficient. That's when the client turned to Paragon for their expertise in automating this personalised marketing process with an AI Recommendation Engine.
The Paragon Solution
We started the project with an analysis and design phase, gathering requirements and refining objectives through remote and in-person communication with key stakeholders across the world. With the information collected, we agreed on an architecture and solution plan, and the development phase began. Our development team, consisting of architects, data engineers, and data scientists, used agile development methods resulting in efficient client communication and trackable performance.
The implemented recommendation engine seamlessly integrated into the client's existing business operations and generated daily personalised recommendations for millions of customers with over 50% success rate. The automated approach enabled the client to target all customers with a personalised list of recommendations, resulting in a six-figure projected annual revenue uplift and a 42% return on analytics investment.
In addition to the monetary value, the client successfully implemented its first AI-fuelled product. We delivered a scalable and multifunctional infrastructure as well as an operational process of how to develop future AI products to the client's tech teams.
By partnering with Paragon, our client was able to streamline their personalised marketing process, resulting in increased revenue and successful implementation of their first AI-powered product.
Key Project Numbers
42%
Project had an estimated 42% annualised ROI from test runs.
6 Months
The project took less than 6 months to implement end-to-end.
40M
Daily personalised recommendations for over 40 million user accounts.