For global jewellery brand Pandora, three pieces of jewellery are sold every second.
The jeweller has 6,800 points of sale across more than 100 countries and 37,000 employees behind it, the brand generated nearly $4.6 billion last year alone.
With such a vast customer base spanning numerous territories, how does a global brand like Pandora ensure that each customer feels valued?
Speaking at Tech Show London, Tiffany Qiao and Mujahid Rashid, senior data engineers at Pandora, explained how the brand used Databricks and AI to address this challenge.
A diamond in the rough
Pandora’s data journey has evolved significantly over the past decade. Rashid explains that the jewellery brand was previously reliant on a business intelligence system, which meant its data processes were slow, siloed, and inefficient.
“Five to ten years ago, we were using a business intelligence (BI) system that didn’t fully leverage our data,” he said. “We had a single team managing data transformations, and our reporting was slow.”
However, over the last five years, Pandora made a decisive shift to the cloud and started using Azure and Databricks to improve its analytics capabilities: “This shift allowed us to gather better insights from various areas, such as store operations, customer behaviour, and our web presence,” Rashid noted.
From there, the company began applying descriptive analytics to understand the past and present and explored advanced analytics to make predictions about the future.
“One of the key achievements has been enabling near-real-time analytics,” he added.
“Our data infrastructure has evolved so that what used to take hours or minutes now happens in seconds.”
Pandora’s infrastructure, powered by Databricks, enables the company to store, process, and analyse data more effectively, moving away from traditional data warehouses.
They adopted a decentralised approach, allowing individual teams the flexibility to manage data in their own way while maintaining consistency and clarity, as Qiao points out.
Dimension 5 Framework
Qiao explained that a crucial milestone in Pandora’s data evolution was the development of the Dimension 5 Framework, a strategic model that outlines five areas that typically focus on leadership accountability, culture, talent management, customer impact, and societal influence.
By addressing these areas collectively, it is believed that organisations can create environments that promote diversity, equity, and innovation while driving business success.
Five ways Burberry is using AI and blockchain to shape luxury retail
Qiao added that this framework provided developers with a streamlined method for defining, testing, and transforming data before it is made available to business teams.
As a result, she maintains that Pandora’s data team can guarantee cleaner, more reliable data, which is essential for decision-making across the business.
Golden opportunity for AI
As Pandora’s infrastructure has become more sophisticated, it was clear that AI would play an important role in transforming its operations, according to Rashid.
He claims that AI’s potential to automate data pipeline optimisation, predict customer trends, and enhance marketing efforts has reshaped the way Pandora engages with its customers.
“AI has helped us fine-tune our pipelines for better efficiency,” Rashid says, highlighting the critical role AI plays in maximising operational performance. The company is also using AI to personalise customer experiences.
Rashid explains that Pandora aims to predict customer preferences based on past purchases and events such as Mother’s Day or Valentine’s Day, or even a customer’s preferences, such as bracelets, necklaces, or whether they prefer gold or silver.
“We want to be able to create personalised shopping experiences for customers both online and in-store,” he notes. “By understanding what a customer has bought or is likely to buy, we can offer them more tailored recommendations.”
“For instance, if a customer shops online, store assistants should be able to access their online shopping history.”
“Ultimately, AI and analytics help us create better customer experiences and drive business outcomes like increased revenue and brand loyalty,” Rashid adds.
“We aim to leverage insights and predictive analytics to stay competitive and deliver value to our customers.”
This also supports their marketing efforts, pushing only relevant content to customers: “This drives higher engagements and reduces the risk of customers unsubscribing from our communications.”
Pandora is also using AI and automation to generate product descriptions for marketing campaigns, reducing development time and ensuring consistency across all product descriptions. “This makes it easier to launch campaigns quickly.”
Moreover, the company is transforming how teams manage data using AI and automation. “Instead of relying on analysts to manually create reports, AI allows the business team to access insights directly,” Rashid said.
“This boosts efficiency and promotes more autonomous decision-making across the organisation.”
How to scale
In terms of business impact, Pandora has been able to scale faster without compromising quality or speed, Rashid says. “Whether launching new products or entering new markets, we’re able to respond quickly because our data infrastructure supports that agility.”
“Pandora has really embraced data as a strategic asset, and this commitment has enabled us to make data-driven decisions, stay competitive, and continuously improve our operations,” adds Rashid.
“We’ve come a long way, from a legacy data infrastructure to a modern, efficient system. We’re confident in our data landscape now, and it’s supporting both our operational goals and customer experience enhancements.”