Collibra improves customer engagement for European DIY retailer Kingfisher

When DIY retailer Kingfisher launched its Hello Casto online virtual assistant, it was the first in the home improvement retail sector to use machine learning (ML) technology to help customers solve their home improvement problems.

Hello Casto, was an example of an enterprise-wide strategy to use advanced technologies like AI and ML to drive tailored customer experiences. Now, most of Kingfisher’s brand websites use an in-house recommendation engine to support customers' buying decisions — fundamental to this is data.

“It became evident to Kingfisher how many of our strategic business initiatives, whether business efficiency or driving customer interaction, depended on data,” said Steve Taberner, Data Governance Manager at Kingfisher. “Specifically, data availability, data quality and putting data into the hands of the people making decisions.”

Kingfisher is a European trade and consumer home improvement retailer with over 1,900 stores and more than 76,000 employees in eight countries across Europe. Its retail banners include Screwfix, B&Q, Castorama, Brico Dépôt, TradePoint and Koçtaş. The business offers home improvement products and services to consumers and trade professionals via e-commerce channels and stores.

“We have millions of customers and understanding how to provide a personalized service to them cannot be done with anything other than data,” stated Taberner. “It is not like the old days when someone walked into a hardware store, and the owner knew them personally. Now, we need data to understand our customers, and their home improvement wants and needs. Personalization is only possible by using data to understand individuals and categorize how we engage with them based on their lifecycle and loyalty journey. Increasingly, data fuels the AI and ML applications we use to enable better customer interactions.”

Kingfisher uses data in a number of ways, such as customer behavioral analysis, managing promotions and campaigns, product innovation, supply-chain optimization and AI and ML applications.

In recognition of the importance of data, Kingfisher has made data-driven customer experiences a key business strategy. It started a governance program with a Group Data Platform and began a sophisticated method of cataloging data in spreadsheets. This involved talking to subject matter experts and system owners to get business metadata such as field definitions, how fields are derived or calculated, data types and data sensitivity. But this effort resulted in 100’s of design spreadsheets, making it difficult to share the knowledge effectively.

Despite being a multi-business, international organization, Kingfisher wanted to be able to leverage its scale and reuse technologies and capabilities consistently. The company needed a commonality of processes and a better way to define and control its data.

Kingfisher looked at all the leading data governance players and partnered with Collibra, preferring an established and proven solution. “Collibra is very high in the Gartner Magic Quadrant for Data & Analytics Governance quadrant and is an industry leader,” explained Taberner. “The solution is very configurable; as a complex organization, we need it to tailor data governance. There are differences across the business and Collibra supports both centralized and localized data governance.”

Kingfisher partnered with Collibra’s onboarding team to implement the cloud-based Collibra Platform, which includes Collibra Data Catalog and Collibra Data Lineage. The platform unified forty siloed glossaries into a single business glossary containing over 4,000 KPIs, terms and definitions, such as how margin is calculated on a sale or product lifetime. This enables business units to access Collibra, understand the group definition of a KPI, and either adopt or localize it as needed.

An in-house API was developed to ingest design spreadsheet data into Collibra. “Ingesting spreadsheet metadata into Collibra has been massively beneficial. Without Collibra, this would lie in 100s of spreadsheets and not be used by anybody. Getting business metadata into Collibra and understanding what a field is and how the business uses it is gold.”

With a common data platform and common descriptions, Kingfisher can propagate solutions across the whole organization quickly and efficiently. “If we took each business unit individually, with different ways of talking about and sorting data, it would take a long time and a lot more cost and effort,” said Taberner. “Collibra enables us to provide a building block framework centrally with localized flavors because not all our businesses are the same.”

Case in point: Taberner and his team worked with Brico Dépôt in Iberia to use Collibra to develop a set of KPIs that followed corporate standards and local needs.

In addition to the spreadsheet API, Collibra integrates with Kingfisher’s Azure Data Platform, and is in the process of integrating with Kingfisher’s Databricks and Power BI. Currently, Collibra is used by 750 analysts but will be rolled out to additional business users in the future. For example, a B&Q store manager in London could use Collibra to look at a report and understand how a value has been calculated.

Kingfisher aims to be a leader in the DIY retail industry for deploying AI and ML to improve customer experience. As it continues to use data to evolve and enhance its customer engagement and business operations, Collibra has become essential to effective data governance.

“Collibra is a key technology at Kingfisher because it helps us use our data for applications that support better customer experiences, guiding them on how to use DIY techniques through to recommending the right products as well as enabling us to understand our customers and their needs better,” stated Taberner. “Data is critical to our business, and there is now a broad understanding that governance is crucial to making it work. Data is one thing, when not controlled and governed, quickly becomes a liability rather than an asset.”