Chris O’Dell, Manager of Planning Statistics at the Department of Planning & Environment NSW shared his insights into building an analytics team at the Data Analytics for Effective Decision Making conference in Sydney this week. The Planning Statistics team is newly formed and was built based on ‘4 Ps’: purpose, position, people and product.
When defining the purpose of your team, Chris urged conference attendees to look at it as though the money invested in your team has been taken directly from frontline staff. Instead of hiring a new nurse or teacher, they’ve invested in a planning and analytical team. To justify this spend, you need to be able to demonstrate your fundamental importance and the value you create. Keep in mind that purpose can change. Make sure to consistently evaluate your purpose and stay adaptable.
Are you meeting the expectations of the Department and your customer? Are you adding value? Chris used Kodak as an example of the importance of positioning. A previous market leader, they decided against investing in the digital camera because it would disrupt the film business. They assumed they knew what customers wanted, and were focused on the past rather than the future. Chris said he found the same thing in the Department. “We were producing products that were great 20 years ago.”
For Chris’s team, he said it needed to go beyond answering questions like, ‘How any houses have you built in Sydney?’ They are now starting to answer questions like, ‘How many medium density homes do we now have, and what are the effects of that?’ In future, they are working towards being able to answer: ‘If we want to continue the development of medium density housing, what should we do next?’
Chris laid out a number of skills that are essential to a successful analytics team. You need a diversity of analytic capabilities. You need a public representative who is going to get out there and sell the value of your team. You need someone who is good at the back end, who understands process. A writer who can build a narrative from your data is crucial, and usually lacking. And someone responsible for visualisation, charts and other representations.
Don’t search for all of these skills in one person – they don’t exist. Support and nurture the skills that people do have, but don’t force them into the areas they’re not comfortable with. “Don’t get DBAs to write and don’t expect a writer to understand SQL code.”
He also highlighted the importance of diversity. People from diverse backgrounds can bring diverse experiences and opinions to your analysis. Someone who grew up in a city will have a very different perspective on liveability to someone who grew up in a rural area.
Before you think about your public-facing products, it’s critical to focus on fixing the back end. This part isn’t fun or glamorous, and your manager is unlikely to take an interest, but you can’t produce the required results until the back end is optimised. To begin with, the Planning Statistics team didn’t have processes, procedures, or a single source of truth. Compiling multiple databases and spreadsheets was essential for successful output.
Selling the importance of this work to a non-technical executive is a challenge. Know your audience. Good content is irrelevant if the delivery is bad. Don’t get bogged down in the specifics – focus on efficiency and savings, not the materials or the engineering.
Missed this event? View our upcoming Public Sector conferences >