denisismagilov - stock.adobe.com
Financial institutions are one of the biggest adopters of big data analytics, but serving up fancy dashboards populated with charts and graphs is not enough, at least at Citi.
Speaking at Tableau Live Asia-Pacific this week, Sarah Burnett, Citi’s head of data democratisation in Asia-Pacific and Europe, Middle East and Africa, stressed the importance of human centric design in data analytics initiatives.
“To design the best [dashboards], we start with pen and paper, not just to draw which chart to use and what the page will look like, but to detail how the story should be told,” she said. “If additional calculations and groupings come up, we can work with our data teams to discuss the best tool for these additional calculations.”
Citi’s human centric approach to data analytics also puts users at the centre of its data visualisations. Burnett said her team of design experts are constantly thinking about how to increase usage of data visualisations and to reduce unnecessary cognitive load on the eye while ensuring the story remains the focal point.
The team has also published style guides to ensure visualisations are consistent, clear and clean, so users can make sense of their data quickly and act on the insights.
“If you’re not thinking human-based design and anticipating the next question that may arise from the visualisation, you can’t blame the tools for poor adoption. Instead, look at how you can improve your design thinking,” Burnett said.
One of the key indicators of success for Burnett’s team is dashboard engagement. Her team strives to understand why users may not be using certain dashboards and where they might be getting the insights from.
“If we are forcing our users to look elsewhere, we have ultimately failed,” Burnett said, adding that the team measures engagements between users and data using log data overlaid with organisational data.
With that data, the team has been able to glean insights on which dashboards are more popular and which ones are not. They can also quickly understand which users are using which dashboards and reach out to them for business inputs.
“We can leverage them to help us promote our dashboards, provide real business insights at community events, and even help us get inputs when we’re planning enhancements,” Burnett added.
Some users may prefer reports to be pushed to them, so Burnett’s team also sends tailored dashboards to users on a subscription basis. “Tracking this usage is also important, as it reflects our engagements,” she said.
Besides serving as an indicator of Citi’s success in data analytics, usage data is also used to determine return on investments, and to reallocate Tableau licences or procure additional ones.
Burnett said the usage data itself is being presented on three dashboards, which are accessible to all users in line with Citi’s mantra of data democratisation.
To engage internal users, Citi’s business and market leads also organise Tableau Days to help users get more out of their data. Activities include live dashboard demonstrations where users learn how they can find insights and ask the experts.
“We also make it interactive, encouraging questions throughout and most importantly, taking stock of the questions that we aren’t answering,” Burnett said.
“As a result of our internal Tableau Days, we see immediate increase in usage in those markets, while getting deeper insights and learning what’s important to our stakeholders. This feedback is key to adjusting our design and future builds.”
Read more about big data analytics in APAC
- Dell Technologies has created a real-time dashboard to make sense of data about the Covid-19 coronavirus and safely guide its employees’ return to the workplace.
- The Singapore Tourism Board has developed an analytics service, among other initiatives, to digitalise the city-state’s embattled tourism industry.
- Australian aged care provider Juniper is a proponent of using data analytics to deliver high-quality care.
- Researchers at the University of Queensland worked with IBM to develop a dashboard with machine-learning capabilities to analyse data for a global study on Covid-19 patients in intensive care.