Bauer Media’s Stuart Forrest: “The Premium Is on Expertise, Not Volume”

The head of audience development at one of Europe’s largest magazine publishers on AI search disruption, why specialist depth beats breadth, and the 47-step briefing tool quietly reshaping how editors

Stuart Forrest is not panicking. Given the scale of disruption currently hitting digital publishing, declining organic referrals, AI overviews cannibalising click traffic, and Google Discover delivering less traffic, that is worth paying attention to.

Forrest joined Bauer Media in 2023 as Head of Audience Development, a role that spans the publisher’s UK and German portfolios. Bauer operates across automotive, women’s lifestyle, entertainment, food, golf, motorcycling and fishing in the UK, with parallel categories in Germany. It is a business built on specialist depth, and that matters more than most people currently give it credit for.

The team of nearly 30 people across both territories spends much of its time not doing things directly, but influencing others to do things differently. “A lot of it is coaching,” he says. “We’re constantly trying to influence editorial teams and product teams. How we communicate internally is something I put a lot of importance on.”

LLM Optimisation Is Not a New Discipline

When Forrest delivered a masterclass to industry peers earlier this month, one framing cut through: LLM optimisation isn’t a new discipline. It’s SEO with a different shape.

“Whatever surface we want our content to land on, the algorithm has to decide whether it’s interesting and valuable enough to show the user. That’s job one. Job two is whether it’s presented in a way that makes them engage with it. Those two principles remain the same.”

The signals that determine whether content surfaces in an LLM — authority, clarity, structured data, topical expertise — have significant overlap with organic search fundamentals. Where it breaks down is economics. Traditional search is built on links and clicks. LLMs synthesise across multiple sources and produce a single answer.

“They might read ten articles and produce one answer. The value exchange for publishers, compared to a referral-based model, is very different.”

Since late January, Google’s AI Overview interface has included a ‘View More’ link that pushes users into full AI mode rather than returning them to blue links. Forrest ran experiments the morning of our conversation. His reading: the AI overview is being deliberately truncated, not complete enough to satisfy the query, but just enough to prompt a click deeper into AI mode.

“If that’s correct, it’s obviously a real concern.” He is straight about the limits of his certainty. “I don’t have the data on this.” What he does note is that citation cards within AI mode have a more prominent format — logo, image, meta description — and whether that translates into meaningful publisher traffic is something he’s actively trying to measure.

The EU press’s view, he says, is that the direction of travel is bad for publishers. “They’re probably right.”

Stuart Forrest, Head of Audience Development

Google Still Needs Publishers. Probably.

Whether Google will continue sending traffic to publishers is, for Forrest, part strategy and part informed bet.

“It knows it needs to find some form of value exchange with publishers so it can have access to that content. We don’t know what that looks like.”

The regulatory context helps. Mounting pressure across Europe and the US over how AI companies compensate content creators is making the asymmetry increasingly visible to policymakers. The argument Forrest finds compelling is that LLMs need computing power, energy, and content. They’re only paying for two.

Google Discover, he suggests, may have functioned as a quiet concession, a way of maintaining referral traffic while traditional search clicks declined. But 2025 has been difficult there, too. News Reach reported traffic down 50% year-on-year. Not everyone is seeing that, he says, but it’s not an isolated case.

Bauer’s organic search traffic is down in the high single to low double digits, depending on where you look in the business. Some brands are still growing in organic year on year, but the composition is shifting. “We’re seeing a change in the types of content that drive those clicks compared to where they were last year.”

His longer view draws on print. People were writing off print a decade ago. It found its level. Search, he argues, has moved from being the great hope for publishers into late-life-cycle territory, requiring active triage rather than denial. Some content types are declining faster than others. The job is to work out which is which and allocate resources accordingly.

The 47-Step Briefing Tool

The most concrete thing Forrest discussed is an internal content briefing tool now in production across Bauer’s UK and German portfolios. It runs 47 steps: extracting data, pulling HTML from competitor pages, analysing ranking positions, and mapping internal link structures. It produces an editorial brief.

It does not write articles. It creates the brief an editor uses to write one. That distinction matters to Forrest.

“We’re leaving all the room for tone, context and expertise. We’re using it to shortcut a lot of the manual research process.”

Where it’s most powerful is in competitor analysis: what are the minimum requirements to compete for a given keyword, and where has a competitor recently outranked Bauer, and why? There are two versions: one for net-new articles, and one triggered by a ranking drop, which reviews what’s now outperforming the existing piece and makes suggestions.

The smallest brands in the portfolio are using it on a self-service basis. Angling Times, which doesn’t have a dedicated SEO resource, is using it to improve buying guides. It tells the editor not which products are best — that knowledge belongs to the journalist — but what the content framework needs to look like to be competitive. “Very often it’s reinforcing their instinct anyway. It’s shortcutting the research process, not replacing the expertise.”

Forrest is direct about the risks of AI adoption in editorial. “We’re in a world where a premium on just creating content no longer exists. The premium is on expertise and authority. We wouldn’t ever want to do anything which dilutes that.” That’s the hill I’d die on in my career.

“The premium is on expertise and authority. We wouldn’t ever want to do anything which dilutes that.”

Bauer Media HQ London Office in Central London Photo copyright Bauer Media

Newsletters: What Job Is This Doing?

Forrest is wary of the word newsletter. Not because email is unimportant — Motorcycle News has a healthy database built on deep audience loyalty — but because the term slides too easily into something indiscriminate.

“You have to be really, really thoughtful about what’s going into that newsletter and why it’s there.”

Empire, Bauer’s film title, should be built around a specific weekly problem: what should I watch this weekend? Today’s Golfer should adjust its email rhythm around tournament schedules. Parkers, the automotive valuations platform, is better characterised as a transactional brand in finance. The expectation isn’t regular weekly engagement. It’s warming up users throughout the purchase cycle so they’re active when they’re next in-market for a car.

“Think about your newsletter subscribers as your most loyal audience. Because if they are, they’ve already read the biggest story from Wednesday. There’s no point putting that in the newsletter.”

Blocking, Measuring and the Value Exchange

Bauer is currently aggressive on AI crawler management. Robots.txt blocking is in place, along with CDN-level restrictions. “You can’t create a market for a good unless you’re controlling access to that good.”

Forrest is tracking four metrics for AI visibility: referral traffic from LLMs (small but monitored), bot crawl data via Tollbit, AIO citation rates, specifically what percentage of ranking queries show an AI overview, and of those, how many cite Bauer, and share-of-voice monitoring, which he treats as directional rather than definitive.

On share-of-voice specifically, he’s sceptical of the tools on the market. Most are sampling data rather than direct measurements. “SEOs are quite comfortable with discomfort around the certainty of data. If there’s an SEO who didn’t do anything until they were absolutely certain, they’d never get anything done.” His approach is to measure the same thing consistently over time and use it as an index rather than a count.

In twelve months, he expects the landscape of content value exchange to have moved significantly. Microsoft’s content marketplace is emerging. ChatGPT has hired a UK-based partnerships lead. The CMA is active. IAB Tech Lab is establishing standards. “We didn’t learn how to trade advertising out of nowhere. Somebody had to write the standards.”

---