Challenge: focus on customer needs to increase retention, profitability and lifetime value

The UK’s top supplier of Liquid Petroleum Gas and part of the global SHV Energy group (turnover: €5 billion), Calor Gas Ltd provides heating, cooking and energy solutions to domestic customers UK-wide and also supplies the heating, catering, agriculture and transport sectors. “With domestic customers, we have typically engaged with every customer in the same way without really taking into account their differing needs,” says Simon Went, Head of IT. “With more than 100,000 domestic customers, it is a big task to change the way we engage with each customer to a more relevant and personal way. To achieve this change we first had to understand our customers better, identify different segments and gain new insights.”

To provide new levels of business insight, the Customer 360 programme was created to better understand customer lifetime value and likelihood of churn, so informing appropriate communications and interactions. This new approach, with a Modern Data Architecture deployed and supported by Datalytyx, has a Microsoft Azure cloud-based data lake that includes Talend technology to bind the entire architecture together and run it day-to-day. Datalytyx had previously provided Calor Gas with Talend data integration, ETL and data warehouse to provide essential business intelligence and management reports.

Why Datalytyx?

Calor Gas’ desire to differentiate directly led to the segmentation programme, to gain the ability to target customers appropriately. “For example, regulations mean our customer contracts cannot be more than two years in length,” Went continues. “As customers approach the end of their contract, we write and ask them to re-sign. Knowing more about what is important to our customer means we have an opportunity to tailor offers which in turn will attract customers to remaining with Calor, i.e. making an offer appropriate to their specific needs or how they want to engage with us. Initially, we used a specialist consultancy to work with our data and create complex algorithms based on three key metrics – churn risk, customer value and segment. That was the ‘data scientist’ part: identifying which attributes were good and which were not so good. Then we needed to industrialise these algorithms and embed them with the relevant customer facing business systems, so the project came to IT, and Datalytyx.”

Data was already being taken from multiple sources through the Talend data warehouse, a previous Datalytyx project. “We could reuse certain approaches and use the knowledge Datalytyx had developed on our business and data,” Went says. “The question was, how do we start our journey towards Customer 360? To mine our data, fishing out insight for different interested parties for both immediate use and more strategic purposes? We already had a Datalytyx managed service for our data warehouse, and Customer 360 is a managed service too. Working with Datalytyx means we can scale up and respond to changing business needs.”

Solution: a data lake founded on Talend, HD Insight and Microsoft Azure technology

Datalytyx advised on a cloud-based data lake using Microsoft Azure and Talend Big Data and master data management (MDM) technology. Datalytyx then set up and deployed a Microsoft Azure HD Insight Data Lake, along with Talend Data Fabric and MDM on Azure also. ‘Operationalising’ the entire process, the new solution aggregates data from multiple systems. “We had used Azure a little, so were already familiar with it,” Went says. “We obviously wanted the cost, scalability and flexibility benefits of a cloud platform. Rather than have multiple cloud providers, Azure made sense. With Microsoft HD Insight, we again relied on Datalytyx to take us through the options: this was our first time using Hadoop distribution to manage operations, scaling and security. We were led by Datalytyx, who gave us confidence about the right technologies and what they could deliver. And so far, so good: it’s all in place, as a production system. The algorithms are used to calculate values on a weekly basis so we’re building up a history, and growing our maturity in how we use the new insight. We’re still learning what factors can move a customer from one segment to another. We’re also now extending the same approach to commercial bulk customers and cylinder customers.”

Benefits: immediate gains, future opportunities

“Before, we knew intuitively we had different types of customers. Now, we have a mechanism to actually engage with customers in more appropriate ways. And we can communicate in the ways they want us to communicate. A ‘bargain hunter’ may only be interested in price. They want a ‘no frills’ service and call only when they want a delivery. Another customer could be more interested in value-added services like guaranteed timed deliveries, telemetry on a tank so they never run out, perhaps having an underground tank installed. This type of customer may be willing to sign up for the added value of a more all-round service. It’s about being in a position to understand and respond to those different requirements.” Calor Gas now has opportunities to centralise all data within its MDA, to support more complex analytics and demand forecasting, and to work towards enabling self-service insights and visualisation by users. “Learning how to use and maximise our data is a journey for Calor Gas – and Datalytyx is a valued partner on that journey,” Went adds. “Datalytyx has the skills and confidence we need: it understands our business, and what we are trying to achieve.”