Why Audience Teams Matter More in the AI Era

Last month’s Journalism AI Festival in London is the gift that keeps on giving. Andrew Deck from Nieman Journalism Lab just published a really good overview of all the various aspects of AI that were discussed. This newsletter is also based on the London conference. Today, we are going to analyze the increasingly important role of audience experts in the newsroom hierarchy. In case you missed my previous post from the conference - here’s my deep dive into The Atlantic’s use of AI.

Today’s sponsor is Beehiiv, which is the platform that hosts this newsletter. Since I migrated News Machines from Substack to Beehiiv 3 months ago, the backend has been working like a charm. Please check out their ad below - every click helps to support this free newsletter. And now let’s dig in.

Dmitry Shishkin has a provocative idea for newsrooms struggling to adapt to AI and changing audience behavior: appoint your next editor-in-chief from the audience department.

Shishkin is a digital transformation expert who spent years running digital operations at BBC World Service, was CEO at Ringier Media International and is now Strategic Editorial Advisor at Ringier Media. His argument: "Your audience team leader, provided they are fully editorially sufficient and editorial sound, has more competencies than your generalist running your business section or economy section."

He made the case at the Digiday Publishing Summit Europe in Lisbon in late October, then elaborated at the Journalism AI Festival in London on November 11. When Digiday called his comments "controversial," Shishkin responded in London: "Show me at least one person who will disagree with what I said.”

Shishkin’s reasoning comes down to career paths. Traditional editors climb one ladder: reporter to producer to assistant editor to senior editor. Audience leaders often start the same way, then hit a ceiling and move sideways into audience work. "At some point, they were going producer, assistant producer, or senior producer, assistant editor. Then they got stuck. Then they moved over to audiences," Shishkin explained at the fireside chat in London. "They started running audience departments. They got exposed to maybe five or six disciplines which [traditional editors] never got exposed to and never had a conversation about."

That exposure matters more in the AI era. When large language models mediate content discovery and AI tools require clean taxonomy to function, understanding how audiences find and consume content shifts from a mere support function to editorial core. 

Modern journalism requires fluency across product, data, commercial strategy, and real-time audience behavior alongside editorial judgment. Audience leaders typically have all of it. Traditional editorial tracks provide depth in one discipline. Audience roles provide breadth across many.

Shishkin's argument gained credence on November 10, when TIME magazine expanded its AI-powered platform beyond Person of the Year content to politics and entertainment articles. The platform now allows readers to translate stories into 13 languages, generate audio summaries of any length, and ask questions about the content through conversational chat.

TIME's expansion represents what Shishkin calls the real frontier of AI in journalism, beyond the "super boring" efficiency gains like transcription and headline generation that most newsrooms celebrate. Multi-modal delivery requires understanding which user needs each format serves.

"You can have an ‘educate me’ piece created as a first-person video, and you can have an ‘update me’ piece written as 200 words, or you can have a ‘give me’ perspective piece delivered as a Q&A," Shishkin said. But all of those outputs require a clean taxonomy so that all the tools interacting with the content can work together, he pointed out. 

Understanding user needs, content formats and performance patterns is precisely what audience teams do. Traditional editors focused on story quality ask, "is this good?" Audience leaders ask, "is this good AND is it in the right format for the right audience at the right time?" That distinction becomes central when every article needs multiple expressions across different modalities.

Shishkin calls this the "golden triangle of content evaluation: topic, user need, format." Newsrooms that don't tag content according to all three dimensions miss opportunities. When LLMs mediate content discovery and multi-modal becomes standard, understanding these patterns shifts from nice-to-have to essential.

Shishkin's strongest evidence comes from moments when audience insights revealed stories editors would never have commissioned on their own.

When the BBC adopted Hearken, a platform that lets audiences submit questions and vote on what they want reported, editors publicly admitted surprise. "I would never have considered writing that article thinking that the audience might be interested in it," one BBC editor said, according to Shishkin. Those audience-suggested articles generated millions of page views.

BBC's user-generated content hub began using Hearken in September 2016, bringing audiences into the editorial process from pitch to publication. The approach worked well enough that in February 2019 BBC World Service launched "We are Bradford", a week-long hyper-local pop-up that let citizens tell their own stories.

Shishkin consulted for one newsroom that created a lot of unneeded stories. The data showed subscribers wanted perspective and education pieces, but they kept getting updates. The editorial team, operating on instinct, failed to match format to audience need.

"You need to climb down from your high horse and let the audience take a decision on what they consider important," Shishkin said. He's careful to clarify this doesn't mean traffic-chasing. 

The distinction matters. Editorial judgment determines the choices. Audience input allocates resources among worthy options. Traditional editors often resist even that limited consultation. Audience leaders, who've been in rooms where data contradicts assumptions, learned humility that pure editorial tracks don't teach.

Then there's timing. "Why am I getting all of the things on TikTok 48 hours earlier than on mainstream media?" Shishkin asked. He remembers the same problem from way back in 2009 at BBC Russian: "I was suggesting stories which I read on Twitter and they only would write about those stories 36 hours later."

Audience teams monitor what people discuss and search for in real time. Traditional editorial teams wait for formal news pegs. In an era when attention spans shrink and platform algorithms favor speed, that 48-hour gap costs reach.

Shishkin points to organizational design as the reason this shift hasn't happened yet. His solution at Ringier was direct: "Rename audience engagement people into editors. You need to have an equal playing field." Words shape power dynamics. Calling someone an "audience engagement manager" instead of "audience editor" determines whether they have editorial authority in meetings.

Rigid job descriptions present another barrier. They make sense in stable industries. In journalism's current state, they create friction. Shishkin implemented a 20 percent rule at Ringier: "Every single senior person needs to be spending 20 percent of their time doing actual newsroom work." Editors who don't upload videos or file stories lose touch with the systems their decisions affect. When he arrived at Ringier, he spent a full day learning how to upload a story to the CMS from start to finish. "I really needed to understand how do I go about all of those things.”

Cultural transformation requires structural change first. Retitle roles, rewrite job descriptions, mandate leadership time on frontline work. Otherwise, audience expertise remains subordinate to traditional editorial hierarchy, regardless of what skills the moment demands.

A January 2025 analysis in A Media Operator also argued that media companies need Chief Audience Officers at the leadership table, noting that typical media leadership teams have CEOs, CFOs, editors-in-chief, and chief revenue officers, but "nowhere in there is the audience represented."

The piece argues for elevating audience leadership to C-suite level, with product teams reporting to the chief audience officer rather than to technology or editorial. "Their job is to deliver a good experience to the audience and maximize the expected behaviors of that audience," the analysis states. "Who defines those expected behaviors? The chief audience officer in conversation with the editorial team and the revenue team."

Shishkin would agree. His point isn't that audience work matters more than editorial work but that modern editor-in-chief roles require competencies traditional editorial tracks don't build. Audience leaders who started in editorial, moved to audience work, and maintained editorial fluency have the full toolkit.