Data underpins UK retailer Marks and Spencer’s ongoing digital transformation. Two years ago, amid a highly competitive market for data skills, the company took matters into its own hands, launching an internal data academy. Now it is taking the academy to the next level, introducing an industry-first course for data scientists. Tech Monitor spoke to Suzanne Howse, head of enterprise data, to learn how and why M&S set up its own training scheme for data, and how it is creating value for the business.
Marks and Spencer’s digital transformation
In 2018, M&S launched a £25m initiative to become a ‘digital first’ retailer. This strategy, which involved an overhaul of its IT operations and the appointment of the company’s first chief digital data officer, comprises various pillars, including ecommerce, the company’s Sparks loyalty card and, crucially, data.
We’re solving business-facing problems, using data science, to deliver millions of pounds of value.
Transforming the way M&S uses data has three components, explains Howse. The first is establishing a new cloud-based data platform, for which the company has partnered with Microsoft. The second is developing the company’s data science capabilities: in 2018, M&S set up a 25-strong data science team, led by its first head of data science. “We’re using that capability to drive value for our customers through our website and through Sparks, and more recently for our colleagues as well,” explains Howse. “[We’re] solving business-facing problems, using data science, to deliver millions of pounds of value.”
The third component is data skills. “We build all of these interesting data science models, but we need to make sure that our colleagues are confident in using them,” says Howse. “Whether I’m a customer assistant or I’m the director of supply chain, I need to have the capability to interact with the tools that we’re building.”
M&S takes a ‘buy and build’ approach to developing data skills, Howse explains. “We like to have a balance of external and internal talent,” she says. “Bringing people in that have lived and breathed in different retailers, or different industries, is fantastic and we like to continually challenge ourselves. But we’ve also always been a company that prides itself on developing our colleagues as well.”
Training up internal staff also means that knowledge of the business is embedded in M&S’s data capability, Howse adds. “If you understand the business and its objectives… it’s a huge win when you’re trying to solve problems using data.”
Furthermore, Howse says, “we have some pretty phenomenal people across the business and we want to keep them. This is another way of showing them that we’re invested in their development and in their talent.”
Data skills training at M&S
Two years ago, therefore, M&S established the BEAM Academy, an internal training programme based on the UK government’s Apprenticeship Levy scheme, which provides funding for on-the-job training, and in partnership with data science and AI education provider Cambridge Spark.
The academy initially offered two courses, at Levels 3 and 4 in the UK government’s grading of apprenticeships. The Level 3 course offers basic skills to any employee who wants to make better use of data in their current roles. Although some participants progress into data specialist jobs, “for the large majority, it’s about being a customer assistant, a buyer or a merchandiser and wanting to be able to use data to help them make decisions.”
The Level 4 course is more technical, and introduces more sophisticated techniques such as data modelling. It is designed for employees with analytical roles, such as finance, Howse explains, although many graduates move into auxiliary data-related roles such as data product manager or data strategist.
Last week, M&S kicked off a new Level 7 course, focused on data science and AI – “a world first in the retail industry,” the company says. The course offers a route for analysts and data specialists to progress to becoming data scientists, Howse explains. It covers the theoretical foundations of data science – how to choose the right algorithm for a given problem, for example – as well as practical experience with tools and how to use them. Ten participants have signed up for the 15-month course, most of whom graduated from the Level 4 course, and will be spending 20% of their time on learning.
M&S worked with Cambridge Spark to design a curriculum tailored to its needs. The courses focus on the tools it uses – PowerBI for data visualisation, Databricks and ML Azure for data science – and trainees work on real business problems. This means they are delivering value throughout their training: some ‘coursework’ projects have delivered six-figures in added value, Howse says.
No isolated data scientist can build anything valuable on their own.
Howse expects many of the Level 7 cohort to join the data science team, while others will continue to work in the business. This will allow them not only to apply data science techniques to their business problems, but also act as a liaison to the central data science team, Howse explains. “No isolated data scientist can build anything valuable on their own,” she says.
The M&S data function has ambitious targets, Howse says, with an expectation to deliver millions of pounds in value to the business. Providing in-house training, she says, will support this ambition in three ways. “We need it to deliver our transformation. We need it to retain our colleagues, and we definitely need it to attract talent.”