Jaguar Land Rover
In February 2021, Thierry Bolloré, chief executive officer of Jaguar Land Rover, unveiled a plan to reimagine the luxury carmaker. The plan is underpinned by connected services, data-driven technologies and what he describes as “radical digitisation of the entire customer journey and ownership experience”.
Given the reimagination of JLR, Clive Benford, head of the corporate analytics programme at Jaguar Land Rover, says a data culture is being driven from the top down. Pointing to the transformation strategy, Benford says: “Agility, digitisation and data are inseparable.”
He believes IT departments need to switch from being application-centric, where they deliver the applications business users require, to data-centric. “If you were to delete all the data from the enterprise, you wouldn’t have a business,” he says. “But you could delete all the applications.”
The reason IT has evolved this way, says Benford, is because systems were traditionally constrained by a computer’s processing power, the level of installed memory and network bandwidth. “Data was a secondary asset.”
Many companies want to be data-driven. In Benford’s experience, all of the value from data analytics is derived by doing things the original systems were not designed to do. “My background from KPMG is to ask what the business drivers are and use data to improve processes,” he says.
One example of using data to improve a business outcome is volume forecasting, he says. “Volume forecasting more accurately means you can respond to market signals sooner. This is worth a fortune. It’s classic improvement consulting.”
Clive Benford, Jaguar Land Rover
Benford says the first time forecasting made a difference at JLR was in 2019 when the market started to turn, and it was able to “slow down vehicle production”.
Another example is in improving how JLR observes signals across its supply chain. “Classically, it is really hard to go from a demand signal to a supply signal,” says Benford. Such a signal may identify that manufacturing will need more parts.
The data necessary to gain transparency across the manufacturing process is distributed across numerous complex data sources from multiple departments, including forecast and supply chain data, parts data from product lifecycle management systems, and car configuration data output by a combination of the car configuration and build simulation systems.
These systems span a diverse array of technology from dedicated mainframes, to dedicated enterprise resource planning (ERP) and material requirements planning (MRP) platforms, to custom distributed car simulation applications. This diverse combination of data meant it was previously impossible to query the data in a timely manner.
JLR uses TigerGraph to combine 12 separate data sources in a graph equivalent to 23 relational tables, spanning the parts supplied by hundreds of suppliers, through the particular model and configurations’ bill of materials, to the manufacturing build sequencing and order forecast for those cars.
Designs to the schema can be made, allowing additional datasets to be added at any time. Data import jobs are generated so the data extract, translate and load process can be repeated as needed. Graph post-processing adds links between the orders and parts for any build date, allowing the query to give outputs across JLR’s entire order book in a few minutes.
Decentralised versus centralised
There are two approaches to building data analytics expertise in an organisation. The first is to hire and train a team to create a centre of excellence and then centrally manage the data analytics for the business. JLR has taken the other approach, which is focused on democratising data analytics.
Discussing the centralised versus decentralised approach to analytics, Benford says: “One of the challenges is that, centrally, you cannot identify everything to do.”
“We are on a 100% path to citizen data scientists. Without a large workforce who are skilled in data analytics, you can only work on a couple of things”
Clive Benford, Jaguar Land Rover
JLR is deploying Tableau, in a bid to democratise data and analytics. Rather than confine data analytics to a central, corporate data analytics team, the goal is to give as many people as possible the skills, tools and confidence to explore and analyse their own data.
“This is a journey, not a destination,” says Benford. Referencing a benchmark from the International Institute for Analytics (IAA), he says, “Amazon has the biggest list of stuff to do,” but, from an analytics perspective, the online retail giant is “the most advanced”.
As Benford notes, nothing happens without people: “We are on a 100% path to citizen data scientists. Without a large workforce who are skilled in data analytics, you can only work on a couple of things.”
In the past year, Benford says JLR has tripled the number of Tableau users. “We have a strong link with Tableau, and the top-down push for decentralised data analytics is huge. Our chief financial officer was the sponsor for a long time. You hear this on quarterly calls, and it cascades down.”
JLR now has more than 1,300 employees actively creating dashboards and analyses, and more than 5,000 exploring this data and finding insights for themselves.
When asked about the rate at which JLR can change to a company driven by data analytics, Benford says that a year ago he would not have anticipated JLR would have a data 101 literacy course: “We chose to build the training ourselves, as it only makes sense in the context of JLR and is relevant to our roles in the business.”
For Benford, access to the data alone is not enough to make a business data-driven. Access to data analytics tools is more important. He believes analytics will become pervasive in office productivity software.
Although such products tend to offer some data tools as part of the suite, often, the request for proposal and procurement process for office productivity tools does not stipulate data analytics. But Benford believes this needs to change as knowledge workers evolve into citizen data scientists.
Read more about data democratisation
- In the enterprise, data democratisation works to break down data silos by opening access to an organisation’s data across teams in an effort to improve workflows.
- Synthetic data generation for machine learning can combat bias and privacy concerns while democratising AI for smaller companies with dataset issues.