Condé’s Commerce Drive, WaPo’s Op-Ed Outsourcing, Metrics Report, and more...
With U.S. publisher Lee Enterprises revealing it spent $2M in restoration costs following February’s ransomware attack, our sponsor this week is timely. Profero is a cybersecurity company specialising in breach readiness and rapid incident response. With its Rapid-IR platform, Profero combines intelligence, discovery, and 20-minute response times into a single subscription model. Protect your media brand—here’s how Profero can help.
This week’s Long Read features a conversation with Markus Karlsson, CEO of Affino—the unified publishing platform powering dozens of media brands. Few are better positioned to take the pulse of today’s publishing landscape.
Let’s start…
Condé Nast Is Transforming Its Brands Into Shopping Destinations
Patrick Gray, Exec Director of Commerce, told the INMA World Congress that Condé drove $600M of sales to its partners in 2024, with affiliate marketing now an upper funnel tool rather than a bottom feeder. Quote: “Our audience knows that when we drop new selections on Vogue, there will be fresh products for them to engage with.” AI has become central to personalising product discovery.
The Washington Post Plans an Influx of Outside Opinion Writers
WaPo aims to expand its opinion pages by partnering with external outlets and platforms like Substack, as well as using an AI writing coach called ‘Ember’ to help non-professional writers craft op-eds. The danger? It could turn WaPo into a glorified contributor platform à la Forbes—something the mischievous staffer who named it Ember is almost certainly hinting at. #gallowshumour
How News Apps Engage Users with Push Notifications
The data is in: An analysis of push notification campaigns sent by 213 news apps shows that push frequency, surprisingly, has no effect on unsubscribe rates or CTR. What works? Early morning campaigns, but keep it light—sports alerts are by far the best performers. The first spike in user interest appears around 7 AM and, rather unexpectedly, Saturday is the best day.
AI Suffers ‘Accuracy Collapse’ When Facing Complex Problems
When AI models hit complex tasks, they don’t just fail—they give up. That's the stark finding from Apple which found that as LRMs neared performance collapse they began “reducing their reasoning effort”, a finding the researchers said was “particularly concerning”. The implication? AI doesn't really reason but just memorises patterns really well.
FREE TOOL: Guardian Shares Free Source Protection Tech
The Guardian has rolled out new secure messaging technology within its main app that lets sources tip it off anonymously. Journalists receive a decryption key to view messages before they self-destruct, disappearing after 14 days. Open source, it’s been (somewhat worryingly) developed in partnership with Cambridge University, traditionally the conveyor belt for MI5. i.e. Check the code with a magnifying glass.
Getty Takes Stability AI to Court in Landmark Gen AI Test Case
Stability, the image-making tool, is in the British High Court for “brazen infringement” of Getty’s photography collection “on a staggering scale”. In what could set a major legal precedent, Getty argues that although licensing creative works is critical to AI’s success, “The problem is when AI companies want to use those works without payment.” Many publishers would agree.
REPORT: Publisher Metrics & Benchmarking Playbook
FT Strategies has released a 2025 ‘Metrics & Benchmarking Playbook’ to help senior teams align on measurement priorities and drill down on the numbers that matter. Covers Traffic & Engagement; Advertising; Subscriptions; Retention; Operational Efficiency; and offers context, benchmarks, and statistical caveats. Ideal bedtime reading…
WTF is AI Slop Doing to Warp Media Metrics?
Staying on the subject of metrics, the sheer scale of AI generated ‘slop sites’ is playing havoc with ad benchmarks. How bad? A 717% increase in low‑quality, AI‑driven domains YoY, with 96,000 simply scraping and monetising clickbait using programmatic ads. Pre-approved domains, clean supply paths, and real editorial signals are more vital than ever.
How Publishers Can Reduce the Risk of Cyberattacks
What do the Guardian, News Corp, Amedia, Lee Enterprises and Kadokawa have in common? They’ve all suffered serious cyberattacks which have had far-reaching consequences. Not every threat can be avoided, but secure backups, MFA, VPNs, credential monitoring, and a solid incident plan can turn disaster into disruption—rather than total devastation. Key takeaway? Preparation.
The FT’s Answer to Tension Between Editorial and Product Teams
The Financial Times has created an ‘Editorial Product Director’ to tackle the disconnect between its editorial and product teams. Clashing priorities, conflicting timelines—and even basic terminology, like what ‘success’ means—have slowed the publisher’s progress. They’ve made product integration a core leadership mandate rather than merely a workflow fix.
SUMMER READ: When The Going Was Good
“You never know you’re in a golden age until it’s gone,” writes Ex-Vanity Fair editor Graydon Carter. This memoir, covering the golden age of magazine publishing, would make even Mad Men’s Don Draper blush. How good did it get? Time Inc. had a 24-hr drinks trolley peddling up and down its Manhattan HQ. Coffee at 11am? Let’s make it a mojito.
AI TOOL #1: Chronicle
Still in beta, but what a tool—Chronicle creates stunning interactive slides and presentations using drag and drop interactive blocks. A user-friendly alternative to PowerPoint, Canva and Pitch, it prioritises design and control over speed—the key USP is that you don’t have to be a designer to use it. At last! Free trial here.
AI TOOL #2: Sana (from Sana Labs)
Sana is a code-free AI platform that plugs into your stack—Google Drive, Notion, Slack, CMS—and automates the grunt work publishers waste hours on. Think onboarding new editors, pulling up old interviews, auto-building compliance checklists, etc. It’s so effective, the Swedish government has just rolled it out to 2.3M civil servants and students.
DATE FOR YOUR DIARY: FIPP World Media Congress 2025
One of the standout events in world publishing, this year’s Congress is set to be another must-attend. Again expertly curated, this year’s line-up includes Jessica Sibley, CEO, TIME; Natalia Gamero, MD, Condé Nast Europe; Jordan Vita, VP Product at The NYT; and many more. Madrid | 21 - 23 Oct 2025
"We’re Past the AI Plateau”: Markus Karlsson on What Publishers Need Next
Markus Karlsson isn’t just another tech CEO with a theory about where media is heading. As the founder and CEO of Affino—a unified publishing platform used by dozens of publishers including The Stage, Inside Housing, and West Coast Today—he has a front-row view of what’s working and what isn’t.
His views aren’t abstract opinions. They’re shaped by the real operational and strategic challenges C-Suite publishers are facing in real time. I asked Markus to share what’s happening on the ground, where the pain points are, and just as importantly where he sees the big opportunities.
While numerous challenges remain, what’s clear from our conversation is that many publishers are now moving from AI-fueled experimentation towards total transformation.
Q: What’s the most urgent issue publishers are facing right now?
Markus Karlsson: Discoverability is a huge issue. Referral traffic from search is collapsing—INMA recently reported that between May 2024 and Feb 2025, traditional search traffic to news sites fell by more than 64M referrals. Publishers are increasingly intermediated by platforms and AI assistants, with ‘zero click’ searches now approaching 60%.
Q: So discovery isn’t just an SEO problem anymore?
Exactly. Traditional SEO tactics are losing effectiveness. Discovery now depends on how well your content is structured, how machine-readable it is, and how easily it can be surfaced by AI.
That’s why we’re talking more about GEO—Generative Engine Optimisation. It’s still early days, but the shift is happening. GEO is not just about content optimisation—it’s about rethinking how content is created and formatted for AI-first discovery environments.
Q: What’s holding publishers back from adapting faster?
Hesitation. Publisher tools are improving, but there's still a lot of second-guessing—especially around AI. There’s a fear of complexity, of giving up control, of investing in something that won’t deliver. But once publishers start using AI tools and seeing results, that changes quickly. It shifts from fear to FOMO. People start asking, “What else can we do with this?” And that mindset change is critical. The opportunities are huge.
Q: Is this just for Affino-built sites and clients?
It’s both. A lot of what we’re doing with AI now can be deployed for all existing media properties, not just ours. We’ve made a point of offering bolt-on modules—like chat interfaces that work with your content, or site-wide search powered by AI. Modularity lowers the barrier to entry and helps build trust.
Many of these AI tools can be integrated into existing sites without overhauling core systems. You don’t need to rebuild your platform to get value. This flexibility has been key to helping publishers get started quickly.
Q: Any recent examples that stand out?
We’re seeing strong engagement from tools like contextual chat, AI-enhanced search, and interactive content assistants. They’re not just novelties—they’re becoming core features that users rely on. Once the value becomes visible, the momentum builds fast—one client was sceptical about an AI chatbot but weeks later was asking for more features.
Q: How does Affino's AI architecture actually work?
At the core are six distinct engines, each focused on a specific function. AI Chat handles user interactions; AI Copilot supports editorial tasks; AI Analyst delivers insights and reporting; AI Agents trigger workflows; Mass Content AI transforms content at scale; Support AI manages common queries. It’s modular because different publishers have different needs. The idea is to let each engine do what it does best rather than overloading a single tool with too many functions.
Q: Are your publisher clients using all six engines?
Not always—and that’s OK. Many start small. Some use just AI Chat or Copilot, then layer on analytics or automation as they grow more confident. Even using a fraction of the tools can have a measurable impact. The real value comes from how well they’re integrated into your everyday workflow. Publishers need to feel in control while still unlocking the efficiency AI offers.
To put it in perspective, even our long-time clients rarely use more than 40% of the platform’s full capabilities. That’s not a flaw—it’s more a reflection of evolving priorities. As publishers get more confident, they build upward and outward.
Q: What makes your approach unique?
We didn’t want to pretend one engine could solve every problem. Specialisation matters. We built six to do specific jobs well. And we didn’t want to bolt AI on top of a legacy stack—we’ve embedded it into the foundation. That changes how publishers interact with their data and their users. It’s about smarter infrastructure, not just smarter features.
Q: How do editorial teams respond to the AI layer?
With caution at first. Editors are rightly protective of brand voice and accuracy. But once they see how AI Copilot helps generate article questions, suggest visuals, or produce quick summaries—without overriding their judgment—the resistance fades. We’re not replacing editors. We’re giving them more time to focus on what matters. Editorial AI works best when it respects the integrity of the team using it.
Q: Final thought—what’s the one thing publishers need to hear right now?
That doing nothing is the biggest risk. The publishers who are trying things—even small things—are already seeing benefits. The ones who are waiting risk becoming invisible. Discovery is changing, and being part of that future means showing up in new ways. The market is moving, and the smart move is to move with it. Huge opportunities await!
This interview is part of our ongoing exploration of how media companies are adapting to AI, platform shifts, and changing audience behaviours. To read more insights from leaders navigating these challenges—or to learn more—visit Affino or follow Markus Karlsson on LinkedIn.
A shout-out to my friend and colleague Monojoy for his invaluable help with this WNIP newsletter. He’s also behind Akibuki, a storytelling-led initiative that tackles content fatigue with scribbles, stories, and a well-aimed nudge at the AI status quo. #thanks