The chief data officer has become one of business’s most critical players as leaders grapple with how to master data. Three experts share the key strategies every business should employ.
Christian Nelissen, chief data and analytics officer at National Australia Bank, says that when he dropped out of uni, “I thought, ‘Well, I’m good at maths and I’d like to have an office one day so banking’s the right thing for me.’” He went back to Curtin University part-time, worked for 12 years at Bankwest and then in consulting in the United Kingdom for more than a decade, by which time he was getting more and more into data. He joined the Royal Bank of Scotland as a data and analytics director in 2010. “The idea that, in my role, I would go and sit at the leadership table and talk about data would have been seen as crazy,” he says. “People would roll their eyes and say, ‘Data is the most boring topic known to man.’”
Maria Milosavljevic became chief data integration officer at the Department of Defence in February. She was planning on becoming a biochemist when she started her Bachelor of Science degree at the University of Tasmania in 1988. “In my second year of uni, I studied artificial intelligence and fell in love with computer science,” she says. “I’ve worked across a range of areas, in academia, industry and government, and my sweet spot is data and analytics applied to security so my current role is a golden opportunity for me. In the 12 years I’ve worked in government, I’ve very much focused on the various elements of data, particularly when I was working in intelligence agencies, where data is the lifeblood of the organisation.”
Kris Matthews is chief product, data and analytics officer at PropTrack Australia, part of the News Corp-owned real estate advertising conglomerate REA Group. “I started a degree in criminal justice but in parallel I was doing an apprenticeship with Ray White real estate,” he says. Matthews got hooked on working in property and as his career progressed, he became interested in the technology side of operations, eventually jumping from real estate to a technology supplier. Ever since, he has worked in data and tech businesses tied to property, including more than 20 years with global data giant CoreLogic. “I have a passion for property and the technology side lit up the opportunity. So that interest morphed into a career.”
Make data part of company culture
All three of these CDOs stressed the importance of fostering a strong data culture led from the top. “You need to build a culture that locks data into the DNA of the organisation,” says Milosavljevic. “Leaders have to be curious – and data literate – and build that mindset. Data is a strategic asset that goes to the heart of an organisation’s mission. Most innovation around the globe relates to the better use of data; it’s key to making an organisation more efficient and effective. You don’t get there unless everybody goes along with that culture of curiosity, as well as having the data literacy they need to achieve their jobs.”
Ditch the data you don’t need
Data storage is cheap today, says Matthews, so businesses have a tendency to amass and hang onto a lot of it. “You can easily sink in data,” he says. “You have to be really clear about the requirements of your data and make sure your data strategy underpins your business strategy.” Then be ruthless if you can. A company like PropTrack, which has to provide timely, accurate data to clients such as banks, needs PII data – Personal Identifiable Information. Holding it requires heightened security measures because a data breach has the potential to expose customers’ personal information. “For businesses that don’t need PII, they shouldn’t capture or store it. It just gives you overheads you don’t need. If you don’t need PII, get it out of the system.”
Hire experts who value clear communication
Leaders must be curious about how data can make a difference in their business and “understand enough to at least ask the question”, says Nelissen. Equally important is that those working in data and analytics – starting with the CDO – be good communicators. “If they can’t explain it to you, you’ve got the wrong person,” he says. “I tell my people, ‘Of course I want you to be capable at analytics but I also want you to be able to walk into a room and explain it to others.’ You need to understand the business well enough that you can convince people to do something different off the back of the analysis.”
Get good data from the beginning
“Data quality must be everyone’s responsibility and it’s something a lot of organisations struggle with,” says Matthews. “When staff members input data from an interaction they have with a customer, be it online or face-to-face, they need to understand the information they’re capturing will be used downstream for a purpose. The quality of the results you get are determined by how accurate the information is at the source.” He says basic training in data collection is a must for all staff. “As to the depth of that training, it depends on the business. It could be incorporated into a course covering the correct use of other platforms, such as the CRM [customer relationship management]. However, businesses that rely heavily on data need a data boot camp for existing staff and as part of induction. This builds foundational knowledge so all employees understand the importance of data to the business and their role in supporting that.”
Strive to make it something you don’t worry about
“I often say my job is to make data like electricity,” says Nelissen. “You don’t spend any time thinking about whether electricity is going to be there or if it’s going to be the right quality. That’s obviously different to where we are at the moment with data.” It might be apocryphal, he says, but in the early days of electricity, there was a role for chief electricity officers, which became obsolete as power became more reliable. “How do I create an infrastructure so data is like electricity? So it’s there when you want it and you spend your time thinking about what you want to do with it.” Ultimately, Nelissen says he’ll be a successful CDO when data is so seamlessly integrated into the operations of a business that he’s no longer needed.
Ask yourself if your data is truly helping your customers
PropTrack is constantly measuring customer engagement around the data it recommends. “We need to make sure it’s relevant and delivers on the promise or value that the consumer is looking for from us,” says Matthews. “Businesses will fail if they don’t understand that and businesses with a really clear view, which measure that engagement, will continue to grow and flourish. Without those foundations in place, you won’t understand if you’re doing what your customers actually want. People will switch off and not take any notice of whatever you’re giving them in terms of recommendations and that obviously damages your brand.”
Don’t create human bots
“There’s a risk that people are so deskilled that they can’t think for themselves,” says Milosavljevic, who likes to refer to David Walliams’ TV sitcom, Little Britain. “The Carol Beer character just sits there saying, ‘Computer says no’, all the time. This is a risk for all organisations. The more we innovate and automate, we have to continually balance things to make sure our people’s skills stay current and we don’t become too reliant on machines.” So it’s not simply about defaulting to the machine’s recommendations but continually upskilling – and discovering nuances in the data.
Work to win the “talent war”
Nelissen says there aren’t enough people to meet the demand for data roles – and more work needs to be done with schools and universities to build the pipeline – but it’s also time to get creative. When he was at the Royal Bank of Scotland, they had a team of 30 people manually reading complaints, a process they automated. “We offered all of them the chance to reskill as data analysts. From memory, 22 of them took the opportunity and then the bank had 22 new data analysts.” He says this mixture of reskilling to help retain people, attracting fresh graduate talent and other “great people” will be part of the CDO’s job until we’re past the skills shortage “hump”.
Don’t use data to do something just because you can
Data science can provide critical insights and improve decision-making under pressure. “Human beings are not capable of processing the scale of the data we’re dealing with today,” says Milosavljevic. It’s taking masses of data points and different types of data, joining them together and making sense of it. “How do we bring that to bear when you’re under high pressure, in a complex environment?” Algorithms can take the data and automate and scale human intelligence. “That’s the art of the possible when it comes to analytics and using it to achieve our objectives. The risk is you can go down all sorts of rabbit warrens. Be crystal clear on the problem you’re trying to solve so you can measure whether or not you’re making a difference.
In research, ‘interesting’ is great but in a commercial or government environment, it must be useful, purposeful and contribute to the organisation’s mission.” Nelissen echoes this. “Data exists to solve business problems,” he says. “There are plenty of cool things you can do with data – and I’ve seen a lot of them – that don’t do anything useful. If you’re not making a difference in the business, you’re basically just amusing yourself. CDOs need to get people thinking and motivated and driven in the right way to solve business problems, not just do the cool things.” In a bank, “data helps us to make better business decisions, from where we should put a branch to how many relationship managers we need in a portfolio.”
Make it clear to the entire workforce where data fits in
“Data needs to be a pillar of your business strategy,” says Matthews. “When it is, you can see a clear link to where it fits into the company’s overarching strategy and how data supports the business.” He says it doesn’t matter where the data sits or what the business structure is, it needs to be obvious “what the business wants to achieve and what function data plays to support that strategy”. A clear data strategy means each employee understands why the information is important.
“For us, it might be, ‘We’re selling a product to a bank that creates revenue for us’, or ‘We’re helping a consumer make a decision about whether they make an offer on this home or not.’ That gives the data teams context, as well as the wider business. And when the data strategy is successful, it drives the business objectives so they are everyone’s objectives. It seems obvious but people underrate the importance of it.”
Respect that data will outlive us
“Data is our most enduring asset,” says Milosavljevic. “In 80 years, the organisation will still be using some of the data we create today. None of us will be here – even the buildings and the IT system will have changed – but the data will endure. So it’s really important that the CDO is long-term in their thinking.”
The roles of the CDO
Maria Milosavljevic sees the CDO as wearing five equally important hats:
Gets the organisation to see the importance of data and have a long-term view of its management.
Most companies operate in silos but data doesn’t – it needs to flow across the organisation. This is about activating data’s latent potential and is often called the democratisation of data, to ensure everybody can use it well.
This is the one that people jump to as the default. It’s about making sure that data is designed, managed and used well. There need to be specific project KPIs around data standards and quality to make sure the data that comes out is fit for the desired outcomes. It’s the same with cybersecurity. It all has to be baked in from the word go. The CDO’s regulator role is about setting in place the right controls to ensure that happens – for the lifetime of the data.
This is about the art of the possible and experimentation but is at odds with the regulator hat, which is about the rules. The scientist is about breaking the rules and pushing beyond the boundaries. The CDO has to balance constraints and controlled experimentation and it’s a constant juggle.
The focus here is scale and industrialisation. To get enduring value we have to industrialise – that’s about connecting the CDO and CIO roles. What’s proven to have value needs to become normal technology that’s embedded in our systems. It’s about applying engineering principles to create something strong, stable and enduring.