Jackie Davies - stock.adobe.com

CDO interview: Robert Michael, group head of data, DFS

‘Data Bob’ is transforming data strategy at furniture retailer DFS, aiming to build a digital twin of the company to improve business modelling – but his first task is to make sure people can trust data

This article can also be found in the Premium Editorial Download: Computer Weekly: Getting comfortable with data

You can tell you’ve made it the top of the competitive world of digital leadership when you become synonymous with your company’s data strategy.

Data Bob – or Robert Michael, as he’s otherwise known – is group head of data at retailer DFS. He was the company’s first data chief and his efforts to turn information into insight have made him a well-respected figure at the furniture brand.

“I didn’t give myself the nickname; it was given to me when I arrived. But it has been a huge benefit in terms of going out and selling the idea and the concept of data management across the organisation. It’s helped to break down some barriers,” he says.

Michael joined DFS Group, which includes the Sofology brand, in February 2020. The chief operating officer (COO) said the organisation had been thinking about appointing a data chief and suggested Michael take a 12-month contract, which would allow both parties to check the suitability of the role. What he discovered on entering the business was information – and lots of it.

“We didn’t suffer from a lack of data,” he says. “In fact, at the time, we probably overdid the data – we definitely struggled from paralysis through analysis. Sometimes you just need simple data to make the right, rapid decisions.”

Michael got stuck into the firm’s data systems and made some recommendations about strategy. Six months into the role, he and his bosses recognised the partnership was working and he was appointed group head of data on a permanent basis.

“It was a good fit,” he says. “I’d made recommendations about what I wanted to do with the teams. I wanted to give people the confidence to use our data to back up what they wanted to do in their everyday jobs.”

Getting comfortable

Michael reports to the COO. Right from the start, he was clear that he didn’t want to be part of the IT organisation. While he works closely with the CIO, who also reports to the COO, the data and technology departments are run as separate entities.

“I took people out of finance and IT to generate one data team,” he says. “I wanted to bring different skilled people from across the business together under a single banner. Some people left, but that was fine, because I could then shape what I wanted the organisation to look like.”

Michael recruited external talent. He also took people from other roles within the organisation who were already seen as data champions: “We can give people the technical skills, but they’ve got that business understanding, and that’s critical – to be able to articulate what the business wants from a data point of view.”

To prove their value to the organisation, Michael and his data team looked for some quick wins. They focused first on finance and a weekly trading report, which was being produced in Excel. This old way of working meant someone would have to come in at five o’clock on a Monday morning and produce the report for the lunchtime trading meeting.

“They’d have to spend seven hours on the report and I didn’t want to be getting up at five o’clock in the morning to do that work,” he says. “But if I’m not prepared to do it, I don’t expect somebody else to do it, either. So, we used technology to pull all the different datasets together using Power BI and started producing that trading report on a daily basis.”

Michael says the organisational impact was instant: “That project proved the value of an integrated approach to data. The initiative brought me praise and gave me the opportunity to look at the data strategy more widely.”

Moving to the cloud

Today, Michael runs a 24-strong data team that includes 12 internal staff and 12 offshore workers. One of his key achievements during three years with DFS has been re-platforming two internal data warehouses – one for finance and another for commercial information – onto Google Cloud Platform (GCP).

The finance data warehouse has already been replatformed and the commercial system will be in the cloud by the end of the year. “That means we’ll benefit from having a single version of the truth,” he says.

“We’ll have a common information hub that sits at the heart of the organisation. As well as going with GCP, I decided to go serverless. I wanted to employ data people to do things with data, not to have to maintain infrastructure.”

“We’ll benefit from having a single version of the truth. We’ll have a common information hub that sits at the heart of the organisation”

Robert Michael, DFS

Michael says using this mix of cloud and serverless has also saved the company money compared to how data was managed historically. When he started at DFS, he discovered that people had been running applications and reports that were no longer required.

“We were able to stop this habit,” he says. “We looked at the reports, we looked at the data, and we’ve actually shown that the data is more accurate in the new system than it was in the old system, so that’s a massive upside as well.”

Turning off old systems is never easy, and a new way of working can lead to pushback from people who are affected by the change. While Michael recognises there were some concerns, his team has managed the switchover effectively and the cultural impact of the digital transformation has been minimised.

“When we turned the old way of reporting off, we had some noise,” he says. “I wouldn’t say we had a huge amount of noise, because we’d done a lot of communication. We’d done a lot of the selling piece up front and we’d actually got to a position where we could make the change relatively quickly.”

Delivering personalisation

Once his single version of the truth is in place, Michael has big plans for enterprise data at DFS. His long-term aim is to create a digital twin that allows people across the different parts of the business to see how potential operational changes might pan out.

“We’ve got an opportunity to start to model everything that happens in the organisation through the various data points if we put everything together in the right way. We can actually reflect what’s happening in the business,” he says.

“We can help shape the business strategy going forwards because we can model the outcomes potentially before we actually go off and do the work. But that’s a massive step change, and it’s about finding the right areas of that particular journey that we can go on.”

Once again, Michael recognises that cultural change will play a key role. Getting people on the same page is the first step in a much longer innovation journey that ends with the development of digital twin technology.

“That project’s about creating trust in the data, which means that I need to pull it together in a consistent way,” he says. “So, if we’ve got a customer demand planning system, and we’re sending out messages to customers and reporting on their requirements, then I want to use a consistent dataset. I don’t want contradictions and arguments.” 

Michael says the end point of all this work on data-led transformation is to provide customers with more transparency and better experiences. While some customers will only want to hear from DFS when their sofa is ready, others will require weekly updates. Michael’s team needs to provide personalised insight for all types of customer requirements.

“A sofa is a big-ticket item in terms of cost,” he says. “If you’re having to wait eight weeks for it to be manufactured, then we need to ensure as a company that our customers are updated when they want to be updated. Every time they call in, we’ve got to have somebody on the end of the phone ready to give them the update that they want to hear.”

Success is about building a relationship and providing insight to customers in a consistent way: “We want to be able to monitor all the various different steps from the sales process through to the delivery process, so we’re able to give them visibility.”

Leading data management

Michael looks back on his career – which has included stints at finance specialist Lowell and facilities management firm Carillion – and says he’s an analyst at heart.

“I like to understand what’s happening,” he says. “If I understand what’s happening, then we can help the business to make better decisions.”

Michael says his career so far has been far from planned; he never set out to be a head of data. However, now he’s in the position, Data Bob is certainly relishing the opportunity to overcome the challenges he uncovers.

“I love data,” he says. “As I say to people when I’m interviewing them, I’ve been at DFS for three years and I haven’t had a boring day. I haven’t necessarily enjoyed every day, but I certainly haven’t been bored.”

Michael spoke with Computer Weekly at a roundtable event organised by consultancy Carruthers and Jackson, which discussed results from the firm’s recently released Data Maturity Index. The survey shows almost a third of organisations (31%) have either unclear data roles and responsibilities, or no formalised data roles and responsibilities at all.

The wide-ranging responsibilities he’s assumed during his career have taught Michael that the chief data officer (CDO) is a key role within business. Not only does every organisation need a head of data, but the importance of information leadership to the enterprise is only going one way – upwards.

“There’s a whole ecosystem that needs to go around data from providing the box for data all the way right through to dealing with security,” he says. “But we’ve got to start selling the importance of data across the organisation. The CDO role is embryonic and it’s growing, and we need to understand how it fits into the organisation going forward.”

Read more interviews with data leaders

Read more on Big data analytics

Data Center
Data Management