How William Hill Pivoted to Real-Time Data and Personalised Customer Experiences

Founded in 1934 as a postal and telephone betting service, William Hill was historically synonymous with the world of horse racing. Since then, the company has continued to grow into other sporting avenues, with the retail arm—which opened in 1966—now amounting to more than 1,400 shops. Fully embracing the potential of the digital revolution, the company now comprises over 10,000 employees over six brands, operating through offices in ten countries.

• Data science teams now able to quickly build new, comprehensive models that address market need
• Easily combine, conform, and analyse data from multiple operational and product systems
• Data science now closely aligned with marketing to identify and improve sub-optimal customer experience

William Hill operates in a tightly regulated industry—a fact that the company takes extremely seriously. Added to that is the complexity of operating in 13 countries, each with their own specific and varying regulations and demands upon compliance.

The business generates vast amounts of data through both POS information from bricks-and-mortar retail sites, and increasingly online operations. Data is part of the fabric of both its business and technology strategy, especially true since a large percentage of its customers moved to digital transactions during COVID-19. The company has increasingly engaged with customers by offering more customised, personal offerings and safer ways to play.

With the business increasingly augmenting traditional bricks-and-mortar offerings with digital ones, this move to the Data cloud has opened huge possibilities in terms of business scalability and faster, more reliable use of data.

According to William Hill’s Global Head of Data Science and Digital, David Parfett: “We wanted to build comprehensive data science models that not only create a better experience for customers, but also enable our ability to quickly enter new markets and create new brands. Agility in the marketplace is key, and this requires a future-proof data platform that allows us to freely scale and innovate.”

Take as an example the “ACCA Club” (dealing with “accumulator” bets), offering customers a real-time bonus balance. Snowflake calculates the balance and pushes back the results of a customer interaction each time the bonus balance is served on a digital device. 

Instead of the traditional batch processing, Snowflake consolidates data from multiple sources, such as laptops and phones. With a comprehensive tech stack working seamlessly through Snowflake, including Tealium and Kafka, William Hill’s data team can focus on near real-time data capture by:

1. Processing data instantly
2. Reacting to customer actions immediately
3. Supporting the business with intraday reporting to make quicker, data-driven decisions

“To create a customer 360 for advanced analytics, you need to be able to combine and conform data from multiple operational and product systems. Snowflake is our centralized data platform that has enabled us to do this, and been a central piece to William Hill becoming a digital-first organisation.”

— David Parfett, Global Head of Data Science and Digital, William Hill

Snowflake retains billions of transaction records, allowing data scientists to leverage both historical and real-time data to generate insights for the business. For example, the data science team now works closely with the marketing team to identify customers that might require follow-on actions, such as steps to improve a suboptimal experience or the offering of incentives to return. 

Migrating from an on-premises solution has required not just overcoming digital hurdles, but also cultivating a different approach and state of mind. It’s not so much a technological challenge as much as a cultural one. 

“It’s not about the tech—it’s about knowing that you use the tech differently. We’re now able to get data into peoples’ hands quickly. If we can do this at speed, then we can then focus resources on work that we know will be valuable.”

— David Parfett, Global Head of Data Science and Digital, William Hill