“Effective newsroom analytics should broaden and deepen understanding of the changing ways readers are finding and engaging with our journalism.”

That’s the definition The New York Times uses for its approach to data insights, and I love the framing and specificity.

During my recent Newsroom Transformation Initiative Master Class, Beyond the dashboard: A deep dive into data analysis and insights, Hayley Arader, executive director of data and insights, newsroom, audience, and storytelling at The Times, stressed that editorial judgment is paramount, aligned with a culture of data fluency.

She delved into attention time as an example of how The Times effectively uses data.

Last year, The Times identified attention time as an important metric:

• The newsroom goal was to deepen reader engagement.

• The business goal was to increase the time readers spend on the platform. 

“A big part of my job is keeping people focused,” Arader said. “People are interested and curious, and it can quickly balloon. We need to explain the why.”

The newsroom does not get financial data, including revenue and subscription information, Arader said. That’s because when the newsroom focuses on producing the best journalism possible, it correlates to their business goals. 

The attention time project allowed the newsroom to cultivate the type of reader behaviour it was seeking in a way simply monitoring pageviews can’t, Arader said.

The attention time metric, to be effective, needed context. Such as:

• Time comparison: Previous time periods, day of week, time of day.

• Relative benchmarking: Topic comparison, section comparison, similarly promoted.

• Segmentation: How do the metrics compare by user type, traffic source, time zone?

• External factors: Seasonality, competitive pressure.

The Times used a model to best understand how story length influences attention time. Longer stories will generally get more attention time, Arader said, but there comes a point of diminishing returns.

“Metrics should be easy to understand, but that doesn’t mean you should be afraid of models,” Arader said. “Really simple models can be a great way to handle context if you frame it well and clearly.”

Their model shows what’s typical as far as attention time for articles of a certain length and then plots articles above or below that average time.

Videos, photos, and interactives also influence time spent. An innovative newsroom needs data that can keep up with multimedia formats, Arader said. Modeling also helped The Times show and track that. 

“It’s so key to being effective in this next chapter of the Internet,” Arader said of understanding how users navigate article pages and multimedia elements.

One of the most important lessons she’s learned in data analysis, Arader said, is the importance of phrasing and how to effectively visually represent the data. The Times was careful not to treat data like a scorecard or to identify stories as over- and under-performing. The Times purposely decided not to use red or green text and to keep it simple with black text. 

The attention time metric sits at the top of the audience metrics dashboard for each story to convey its importance, and everyone in the newsroom has access to this page.

“We want to keep it simple and make it about learning and incentivising the types of journalism we want to produce here,” Arader said. “We created and circulated materials to explain the metric to everyone and created a great FAQ.”

Finally, Arader said, the attention time focus worked because it came from the bottom up and the top down. The Times’ managing editor e-mailed the newsroom to explain the importance of focusing on attention time, and audience editors work directly with journalists on it.

“Our newsroom is data-informed, not data-driven. We say this all the time,” Arader said. “Editorial judgment comes first, and that’s what has helped us be so successful with data.”

If you’d like to subscribe to my bi-weekly newsletter, INMA members can do so here.

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