As chief data officer at Sainsbury's, Andy Day helps the retail giant's 190,000 employees make the most of the huge amounts of data it collect about 40,000 products across 1,300 stores.
Day joined Sainsbury's in August 2016 after two-and-a-half years as CDO at publishing firm News UK. He spoke with ZDNet about how CDOs can help encourage data-led transformation in businesses.
1. Get stuck into organisational change
Day spent most of his first two months in the job talking with colleagues. He met with more than 200 people and wrote 650 pages of notes. Day advises other CDOs who are new to their roles to take a similar approach. "Spend time listening to people," he says. "Make yourself available and suck the knowledge from experienced people."
This information helped Day build his strategy for data-led change. He says his two major achievements since joining the firm have centred on the data environment and the development of in-house skills.
A new organisational and operational model has been signed off and Day continues to recruit key talent. He has created an analytics centre of excellence, with 60 members of staff working on a priority list of applications. "We've pulled together a pretty impressive leadership team," he said, speaking to ZDNet at the recent Data IQ Summit 2017 in London.
"We've got a lot of talent in the team and bringing that capability together represents a big win for the organisation. Our approach has been about corralling the talent we already have and augmenting that with people from outside. I think that grip on capability stands us in good stead when it comes to working on our major data projects as we move forwards."
2. Look to innovate at the edge
Day says CDOs must keep one eye on the future and has created a data lab function, with a mandate to "innovate at the edge".
Day runs a small team of cross-functional specialists, and he advises peers to create a dedicated innovation team. This department should provide a safe environment for people to innovate, where individuals have constant links back to the wider organisation, but where they do not have to deal with the pressure of day-to-day business concerns.
Projects have already been launched in certain areas of the organisation and Day has bigger plans for change. "So, for example, can we talk to farmers about how we instrument their farms, so we get better data, so that we help them breed better cows or grow better lettuces? Or can we put sensors on the side of our fridges, so that we can predict break-downs and provide better continuity of service?"
Some of these ideas, says Days, might lead to a dead end. But the innovation team has the mandate to try things. "Their objectives are to come up with a set number of innovation projects each quarter and to scale a certain percentage of those so that they create value," he says.
3. Grow the seeds for change across the business
CDOs, therefore, most find ways to turn creativity into a business return. Day is now identifying what he refers to as his 'moon shots', which are the big - potentially game-changing - ideas that could create significant value for the business. He is building teams around those ideas and creating programmes of work.
Progress has already been made. Day refers, for example, to the firm's data lake programme. Six months after starting the initiative, the firm can process 300 transactions per second from tilling systems. The information is now being used by data scientists across Sainsbury's.
"I want to create a data environment that's properly fit for purpose," says Day. "Most importantly, I want to deliver projects that provide fantastic customer experience and significant value to the business. We're trying to make everything better through analytics and, if you set that seed in everyone's minds, your work as an executive becomes easier."
4. Sell your brand to help overcome skills concerns
Day says the key challenge is finding great data talent. His aim is to build a world class, end-to-end team with strong capabilities across data management, governance, analytics and reporting. Some of that capability will be based around nurturing internal talent and some will involve bringing in external people.
"In the high-end data analytics space, we are - as an organisation - competing with cool brands, such as Google, Facebook and Amazon," he says. "Everywhere you look there's cool brands doing great things with data. So, there's always specific companies that can help you develop your career as a data scientist."
Day says Sainsbury's, despite the high demand for data talent, is trying to tell a similar story. "There's a bit of everything at our firm, whether you're the expert on pricing, gamification, logistics, or agricultural science," he says, before suggesting CDOs competing for talent must sell the benefits of their brand.
"You can be a specialist, but you can also learn a general set of skills. When you get inside a big business, such as Sainsbury's, that's what really brings data to life. We're aiming to turn a famous brand into a cool place for data scientists to work."
5. Keep the board aware of your great work
Days says one key lesson sits above everything else: be ruthlessly focused on outcomes. "You must have a business problem that you're trying to solve, as opposed to thinking that by building a fantastically technical data architecture, with the best data scientists, you will suddenly be able to deliver millions of pounds of value," he says.
Success requires constant engagement, says Day. In its simplest terms, his job boils down to two simple components - first, helping people to learn the art of the possible and, second, working out how people across the rest of the business can make the most of data.
Day also recognises that success remains a continuous work in process. He says his ability to deliver great results means demonstrating regular proof points to the board. "If the chairman and chief executive can draw a direct line between the investment in data and the bottom line and customer satisfaction, then that's the be all and end all," says Day.
Read more about technology leadership