Thousands of fans flooded the streets of Liverpool this past weekend to celebrate the Champions League victory of their home team.

Unknown to most fans, one of the key changes in the club’s strategy in the last few years was its use of data, particularly in recruiting players. It used data on the progress of 100,000 players across the world to recommend to managers who they should recruit and how the players should be used. These decisions are typically based on how good a player ‘looks’, rather than on their actual performance over previous games.

In government, like in football, leaders must “move away from intuition, gut feel, and experience to relying on data to tell them what’s happening and what actions they should be taking,” says Leslie Ong, Country Manager, South East Asia, at Tableau, a leading analytics platform company. Easier said than done; here’s how leaders can change the culture in their agencies to be more data-driven.

1. Create a common language

If data-driven decisions are here to stay, employees need to be comfortable to read, interpret, and communicate with each other using data. Leaders must create a common language around using data.

While not everyone needs to learn about tools, software and programming languages, there are fundamental skills like knowing the types of data available; how they can be manipulated; and asking critical questions that are crucial to have a data-driven culture.

Ong reveals that at Tableau, every employee goes through a data literacy course as part of their onboarding process. He explains that without understanding the language of data, businesses will not see the results even after providing everyone access to analytics tools.

The New Zealand Prime Minister’s new wellbeing budget is a fantastic example of leaders using the language of data to change culture and mindset. The government will measure progress on five key social outcomes, such as mental health and child poverty, and use this data to make future spending decisions. Ministers have been tasked to find ways to work together to meet the five goals and align their own portfolios using the data.

2. Empower your teams

The second step is to empower people who work in your agency to use data to make decisions whenever they need it.

This requires finding the right balance between freedom and control in the way data is governed: too much flexibility opens up the risks of using incorrect data or unauthorised access to data. And too rigid controls will mean that people will not follow the rules and find their own ways to work around it, which limits collaboration across departments. “You have a situation where the data is not available to everyone, or where is available, it might not be the single source of truth,” says Ong.

Leaders can overcome these challenges by providing access to easy-to-use tools, along with data. “It is about putting the right kind of tools in the hands of people. So you need to give them something that’s easy to use, intuitive, such that it really helps them with the work that they need to do.”

The Singapore Government, for instance, has partnered with Tableau to train 1,500 public officers in how to understand and use data. This training, together with access to analytics tools and data, is helping civil servants ask the right questions and solve complex problems.

3. Build communities of practice

Children learn best through play, and it’s no different for adults. Leaders need to provide time and safe spaces to build informal communities for people to experiment with and become familiar with the new culture. This will bring to the fore people who are keen to champion the use of data analytics across the organisation, allowing leaders to build allies to create change.

One way to do this is to carve out a physical space where anyone with an idea to play with data is encouraged to try it out with others with the same vision. It creates an environment where people feel safe to fail, learn from experts in other departments, and come up with data-driven innovations together. “A lot of organisations that we’ve seen that have been really successful in driving a data culture, have been really good at building internal communities,” Ong says.

One of these is the Ministry of Manpower in Singapore. While ministry has its own dedicated data analytics team, it also creates informal groups and spaces for anyone – including senior management – to give data a try. The data unit also helps ministry officials test new ideas, during which time their regular work targets are suspended so that they can try new things without fear of failure. “We’re trying to nurture a habit of testing and experimentation,” says Hefen Wong, who set the unit up with a vision for “all decisions to be data-driven”.

Leaders can use these three elements to build teams that make better decisions with data. Else, as statistician W. Edwards Deming said: “Without data, you’re just another person with an opinion.”