Two years ago, if you’d asked topic editors at The Sydney Morning Herald and The Age who compiled the data in front of them, very few would have been able to answer. The teams were on separate floors and there was little interaction between the data engineers, analysts, and broader newsroom.

Slowly, we’ve managed to bridge that gap. We had to explain the intricacies of newsroom workflows to engineers and analysts, and translate data into actionable insights for editors. The result has been a better equipped newsroom, better informed data team, and better served audience.

The data team is more integrated with the newsroom than ever before. We invite them to regular editorial meetings — even if it’s “just in case” someone asks a question about how the data works. The best part is, they genuinely find it interesting to see how the broader newsroom responds to the work they’ve done. It gives them greater insight into how we use the tools they give us and how they can improve them.

We’ve found the more the newsroom can put a face to the team behind the dashboards they use, the more likely they are to reach out when they spot a bug, find something interesting in the data, or make a suggestion to enhance the dashboard experience for everyone.

Slack channels do most of the heavy lifting for us, but we also have regular catch-ups. One of these is for the prioritisation of ad-hoc analysis, and another is for sharing insights from that analysis back to key editorial stakeholders. From there, we discuss the best way to share this with the broader newsroom.

Usually this happens in a fortnightly “audience insights” session the audience team hosts. It’s scheduled for lunch time, is open to the entire newsroom, and staff will dip in and out depending on their workload for the day. There are slides to show key takeaways, but it’s very much a casual conversation where an insight is shared and the newsroom jumps in with more questions or observations.

In these sessions, we cover everything from content read most in conversion journeys to which newsletter readers are the most engaged. We also use them as an opportunity to showcase work from around the business. We’ve had our product director share their team’s broader product roadmap, tying each item to the insights we’ve seen in the data. We’ve had the head of consumer subscriptions share brand campaigns, research, and the latest acquisition numbers and churn rates.

Unsurprisingly, sharing insights leads to more questions. If we can’t answer them on the spot, the data team ensures those questions are answered in their next batch of analysis. And on the cycle goes.

We always talk about building trust with our readers, but this feedback loop helps build trust between teams. The constant communication between product, marketing, and editorial means we’re more aligned than ever before. This presents advantages for our audience: The content we produce is reflective of the brand campaign they see in market while the product experience they have with our platforms complements our editorial strategy.

We’re constantly looking for ways to improve the relationship. This isn’t to say we’re never satisfied. We’re just conscious of complacency. In the same way the newsroom uses dashboards to see how our readers are engaging with content, our data team tracks how we’re interacting with their dashboards. Very meta, I know. They come to us when they notice a dip in usage of specific dashboards. It forces us to question whether we can improve that dashboard or why it failed where others have succeeded.

I’ve written before about how our newsrooms are focusing on what content doesn’t work to create more of what does. The data team follows the same principle. It’s led to us retiring some dashboards, like the monster-sized analysis dash that tried to do everything for everyone, and focusing on dashboards that answer questions tied to specific goals. This process keeps us focused. Instead of the newsroom asking “can we have XYZ metric added” and jumping straight into solution mode, we question what we’re trying to achieve and work on the best way to measure that goal.

We’re seeing results. Our Topic Editor Dashboard (affectionately known as TED) was born out of conversations with editors.

One of the key objectives when developing TED was to understand what content wasn’t working so that we could improve it or stop doing it. We needed to do less, but make it better.

Editors were looped into the development process early. The data team built a prototype, and a small working group provided feedback until we felt the tool was intuitive enough for broader rollout. Someone from the data team attended every training session, and each of those sessions provided clues about how to improve TED to meet the needs of the people using it.

In the past year, we’ve reduced the total number of articles published by 19% and increased subscriber pageviews per article by 12%. This process was such a success that we’ve mimicked it for other dashboard creation, like our newsletter dashboards.

Not only is our newsroom engaging more with data, but our audience is engaging more with our content.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .