The media landscape is radically transforming, with AI at the helm. This evolution sparks heated discussions on AI’s role in enhancing journalistic endeavours, fears of job displacement, and the overarching question of how technology will shape the future of news.

Amidst this whirlwind of change, innovative collaboration is taking root at Advance Local in the United States, where humans, AI, and robots are joining forces to redefine journalism. 

During last week’s INMA Media Subscriptions Summit —  sponsored by AdvantageCS, Chargebee, Chartbeat, Google, FT Strategies, Piano, United Robots, WordPress VIP, and Zuora — Lamar Graham, vice president of content strategy at Advance Local, and Matt Dowling, director of breaking and local news at NJ Advance Media, provided insights into their journey of deploying and scaling automated content across their platforms.

When robots join the newsroom

The family-owned Advance Local brand has a strong newspaper heritage and operates across 10 U.S. markets, including New Jersey, Massachusetts, and Michigan. Its inspiration for automation came from the Springfield Republican in Massachusetts, which harnessed data to publish weekly listings of home sales in Hampden County — boosting subscription sales significantly.

Recognising the potential to replicate this success on a larger scale, Advance Local envisioned a strategy to automate data collection and publication, aiming to enhance subscription sales without the labour-intensive process of manual data handling. 

This vision led to a partnership with United Robots, a Swedish company specialising in automated editorial content. Since 2015, United Robots has delivered data-driven content to over 200 newsrooms, amounting to five to six million published pieces.

Cecilia Campbell, senior strategy advisor at United Robots, emphasised the importance of reliable data as the cornerstone of their technology. As she explained, their process begins with data analysis, which then informs the storytelling.

“The data points are the facts of the text. There are no other facts going into the text,” Campbell explained. “So you know you can rely on it being correct as long as the data is correct.”

One notable distinction in its approach is the use of rules-based AI, contrasted with the more recent generative AI. Campbell noted the necessity of human involvement in the initial stages of rules-based AI, where programming and text segments are crafted by humans, ensuring precision and reliability that GenAI currently cannot match without extensive editorial checks.

“There has to be editorial checks at the end. So you lose the volume and the automation,” Campbell said.

Weather alerts, home sales on autopilot

Advance Local’s pilot projects, which automate weather alerts and real estate updates, underscore the tangible advantages of integrating technology into news reporting. This strategic automation enables instantaneous weather updates, leveraging real-time data from meteorological services to ensure accuracy and timeliness in weather reporting.

 “United Robots is scanning the weather service literally every one minute and pulling the latest data, so we’re always up to speed,” Graham said.

Automation has revolutionised content creation for the real estate section, offering a sustainable model for expanding coverage without incurring significant costs. This approach streamlines the production of home sales reports and allows journalists to allocate more time to investigative stories and in-depth analysis, where human expertise is irreplaceable. 

“We would not be doing this content if it weren’t for automation. There’s no way we’d be gathering all this stuff,” Graham said.

Moreover, automation presents an opportunity to delve into hyper-local journalism, a sector that has declined in traditional coverage over the years. Graham sees this as a chance to reclaim and enrich local news reporting. 

“We don’t cover local news the way we used to 20 years ago. And this could give us a hyper-local footprint in our markets with valuable content that people are really interested in.”

A culture of co-creation

 The positive outcomes of Advance Local’s innovative approach are noteworthy. Implementing automated content has led to a significant uptick in engagement within the real estate sector across various markets.

In New Jersey, there was a remarkable 137% increase in the number of people engaged with the real estate segment. Michigan experienced an even more substantial rise, with a 174% increase, and Oregon saw a 68% increase. Furthermore, Advance Local’s real estate content has gained increased visibility on Google Discover, which serves as its primary search platform.

These metrics underscore the efficacy of integrating automated content into their strategy and highlight the potential for scalable, data-driven journalism to attract and retain audience interest in specialised content areas.

Building on these positive outcomes, it’s crucial to navigate the challenges that accompany the integration of automated content into journalism.

While the significant increases in engagement and visibility on platforms like Google Discover affirm the value and potential of automation in enhancing content reach and relevance, concerns around reliability, transparency, and search equity pose complex questions for the future.

As Advance Local continues to pioneer this blend of human and machine collaboration, the lessons learned and the strategies developed will not only shape their own operations but also offer valuable insights for the broader media industry.

This journey underscores a pivotal moment in journalism, where the balance between technological innovation and maintaining journalistic integrity becomes a blueprint for future endeavours in the digital age.

“This is real culture change, machines and humans working together,” Graham said. “There’s a lot to learn about this. And this is kind of a low-impact way to get us into an environment where eventually we’ll be using more generative AI.” 

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