TL;DR:
- Harrods, the iconic luxury store, revamped its loyalty program by harnessing data and AI.
- Carys Lees, Head of Data Strategy, shared the journey at Big Data LDN.
- They consolidated data from various sources, enabling a holistic view of customer behavior.
- A data platform powered by Microsoft Azure was adopted for seamless integration.
- Automation and ML reduced the need for human intervention in monitoring customer behavior.
- Harrods is now using data to optimize clothing orders, ensuring better stock management.
Main AI News:
Harrods, the renowned luxury department store celebrating its 175th anniversary next year, has undergone remarkable transformations in its century-long existence. Since 2002, the establishment has leveraged a loyalty card program to bolster sales and customer retention, a strategy akin to other reward initiatives. This program allows patrons to accrue points on their purchases, redeemable for rewards that grow in value commensurate with their spending. One noteworthy evolution in recent years involves the innovative utilization of purchase data to gain deeper insights into customers’ preferences.
Carys Lees, the Head of Data Strategy at Harrods, shared this strategic shift during her presentation at Big Data LDN in September. In 2021, amidst the turbulence of the Covid-19 pandemic, Brexit, and supply chain disruptions stemming from panic buying, Harrods seized the opportunity to embrace artificial intelligence (AI) and machine learning (ML). “These challenges primarily impacted supermarkets but had ripple effects,” Lees explained. “It was a chaotic period, but also an opportune moment to venture into AI and machine learning.”
The pivotal first step involved consolidating available data from myriad sources into a comprehensive data lakehouse. As Lees elucidated, “Previously, if you purchased a Prada tote bag and subsequently booked a haircut on the fifth floor, we wouldn’t have connected the dots to identify you as the same customer. We successfully rationalized this data to create a unified view of our customers, enabling us to discern trends and brand affinities.“
Harrods’ already highly-penetrative loyalty program played a pivotal role in the success of this endeavor, facilitating the formulation of a data strategy and attracting key stakeholders and investments.
The critical trifecta of people, processes, and data followed. “Once we secured the data, we needed an effective platform,” Lees continued. Harrods selected a data platform powered by Microsoft Azure, provided by Ascent. Building this platform necessitated seamless integration with the store’s existing infrastructure and IT team. Lees added, “Subsequently, we focused on empowering our data science personnel to transition to this cloud stack, resulting in a palpable enthusiasm within the organization.”
Subsequent phases focused on deployment, harnessing automation and ML to minimize ongoing team involvement. Lees emphasized, “Our objective is to free our team from babysitting outdated technology. When anomalies in customer behavior patterns arise, we swiftly detect and address them, training the model and reassessing the situation. Human intervention becomes necessary only if the re-evaluated model veers off course. This approach allows our teams to concentrate on value-added tasks rather than technical minutiae.“
Leveraging Customer Insights for Profitable Growth
Presently, the team is engaged in a project aimed at streamlining clothing orders in various sizes. Lees explained, “We operate on a two-season calendar, each spanning six months. Managing stock in diverse sizes entails significant financial implications. Failure to meet demand due to inadequate stock or excessive terminal stock can be costly.” The team’s solution involved data engineering efforts to streamline sizes into extra small, small, medium, large, and extra-large categories. Lees highlighted, “Expanding our inventory in the extra small and extra-large categories resulted in a broader size distribution, increasing the likelihood of customers finding what they seek. While the concept is straightforward, the key lies in effectively communicating these insights to our buying team.”
Lees emphasized that the primary objective of this project is to stimulate new business and enhance profitability for Harrods as a whole. “The critical enablers include the strategic placement of the right data, an apt platform, a robust loyalty scheme, and a team of skilled individuals adept at experimenting and analyzing data. However, what binds it all together is the power of automation.”
Conclusion:
Harrods’ data-driven approach to enhancing customer insights and loyalty program effectiveness demonstrates its commitment to staying at the forefront of the luxury retail market. By embracing advanced technologies and fostering a culture of data-driven decision-making, Harrods is poised to continue building strong customer relationships and driving sustainable growth in a highly competitive industry.
