Firms need to think strategically when building up their data teams, according to data leaders from two South African banks.
According to Esther Munyi, chief data and analytics officer at South African-based Sasfin Bank, many companies tend to build teams without really understanding what that team is going to deliver.
Speaking on a panel at Africa Tech Festival earlier this month, Munyi added: “Businesses must think about the initiatives they are trying to learn, and then, based off that you can hire the right people.”
Businesses also need to assess the organisational model of their data team, she added. “Most organisations build a central data team that works on multiple projects or multiple business functions, and that creates consistency — it reduces the chance of key-person dependency.
“The other option is decentralised teams or teams that are embedded in the business units, the key focus is on the business units’ requirements, not necessarily at an organisational level,” she explained.
The third option, which Sasfin Bank has adopted, is a semi federated model, or a ‘hub and spoke model’.
“It means you have certain data teams that are embedded in business, and you also have a central team that works on organisational -evel or enterprise-level projects. That way you’re creating consistency around the standards around how data is being used and managed,” she said.
It’s a bit like a sport…
Nollie Maoto, chief data and analytics officer, Merchants, First National Bank (FNB) South Africa, said to think of a data team like a sports team.
“Every sports team has positions. Those positions are clear about what they can and can’t do, what boundaries they can and can’t play in. So, it’s having clear roles and responsibilities, but more importantly, understanding the sport and what you’re trying to achieve.”
In terms of skills, third panel member, Kamogelo Radebe, senior data manager at FNB, said that change management implementation is becoming more significant than technical skills in data strategy.
“So, what I do with my team is ensure that part of their training and personal development includes an element of leadership skills because to be able to lead other people doesn’t always come naturally for technical people,” she said.
Levelling up talent
According to Munyi, organisations should take a risk-based approach when upskilling in-house talent. Take someone that might have a data-related skill — a financial analyst, a risk analyst, and then invest in training and upskilling them to learn that role.
“When you take risks on people, and people know that you’re taking that risk, they tend to deliver they tend to want to impress and want to make sure that that they will move ahead,” she said.
To get an all-rounded data expert, Munyi added that “individuals require softer skills” such as telling the story behind the numbers, particularly in a way that the wider business can understand.
Munyi also said that businesses should look to invest in academies to retrain existing employees.
“So, we’ve created a Data Science Academy… long story short, that’s actually how we’ve created a lot of our data dream team.”
But being data-driven doesn’t happen overnight, Munyi added. It requires a cultural transformation and incorporating a change agent or change manager will ensure the initiatives that you are building succeed, she concluded.
To read more stories on big data click here
