Can you say you’ve had a busy week when everyone has had a busy week? Probably not but I’ve really enjoyed talking to publishers and brands in the information sector about their brand marketing challenges when it comes to those pesky answer engines. If you want to chat about this, please reach out, otherwise I will see you at Newsrewired in London.
Let’s leap in.
The Three-Layer Cake: Public-Interest Information in an AI-Intermediated Age
Gina Chua and Adiel Kaplan at CUNY’s Tow-Knight Center spent a year talking to some very clever people about the impact of AI on newsrooms. The result is a report that doesn’t pull any punches. They say we are undergoing ‘a fundamental reordering of societal information infrastructure’ and entering a ‘new information economy.’ We all know about the attention economy but now we must adapt to the ‘intention economy’ where ‘the new opportunity is with systems organised around what a person is trying to find out - to fulfil the individual need behind a question, explicit or implied.’
There’s a big focus here on an audience who are now consuming news via LLMs where the answers are highly questionable. Newsrooms must adapt by also working together, along with the wider information industry, on standards and protocols to establish a ‘traceability framework’ so readers (and journalists) can find/signpost sources more explicitly.
I really liked how this paper highlighted this is an information industry problem, not just news. From the work I’ve done with trusted information providers, I can confirm we are all facing the same disinformation challenges. The time to poo-poo collaboration is over.
The Verge switches up its audience team to prepare for Google Zero
There’s a lot going on in this article. It’s always interesting to read about how a newsroom is adapting but I would also encourage you to read Barry Adams’ post on this subject as well. What do you think?
Anita Zielina: Is your product strategy and roadmap built for the future of media or only for efficiency?
Anita Zielina and Aron Pilhofer ran an invite-only roundtable for chief product officers at the News Product Alliance Summit in Chicago, and they shared the slides. Her starting point is that AI is a question of strategic positioning, and leadership teams answer it through what they choose to build, buy, partner on and stop. Her own view is blunt: too much of the average roadmap goes on internal efficiency and not enough on the readers we’re building for.
Again, a really useful summary of current newsroom challenges.
Dmitry Shishkin: Why content strategy should start with user needs
I think I’m starting to sound like the Dmitry Shishkin fan club as I featured him last week - but he is a great communicator and this is another reason why you should invest more time to user needs.
Are the weekends your golden Google Discover opportunity?
Newsifier found that, outside sport, an article published on Saturday or Sunday gets 54% more Google Discover traffic and 21% more Search traffic than one from the same publisher on a weekday. The reason is odd: newsrooms publish about half as much at weekends, but Discover keeps filling the feed. The report has five more findings on headlines, including that naming two things (the player and the club) lifts Discover traffic by 12%, and a money figure lifts it by 24%. Always worth scanning findings like these!
Making content from Discoverable Servers easier to discover
Big news this. Discord is opening up its biggest public communities to web search. From 16 November, every Discoverable Server (public servers with at least 1,000 members, and there are more than 63,000 of them) gets a Server Web Page, which may show automated summaries of what people are talking about. From mid-December, individual answers from public channels can appear in search results as well, without usernames or avatars. Admins can opt channels out and members can opt out their own messages. In the UK and EU, members are opted out by default.
Why it matters: Google already leans on Reddit for answers from real people. Discord could become the next forum it draws on, especially in gaming, so gaming publishers may face more competition for help and guide queries.
How to Remove Bot Traffic from GA4 With a GPU Check
This is a really handy walkthrough about how to filter out bot traffic from your reporting. Crucial because over-inflated traffic numbers is something publishers really don’t need right now.
YouTube’s new search feature takes a page from Google Search, and that’s bad news for creators
Oh great. YouTube is testing ‘Ask YouTube’ AI answers at the top of desktop search results, for some signed-in US users searching in English. They show up for how-to and open-ended questions, such as how to get marks off painted walls. The card mixes written advice with timestamped moments from relevant videos, and you can ask follow-up questions or scroll past to the normal results. It follows the launch on mobile. YouTube hasn’t said how creators will be paid if fewer people click through to their videos.
Why it matters: YouTube search is heading the same way as Google’s. If you publish how-to or explainer video, expect the same squeeze on clicks that AI Overviews put on web traffic.
AI Visibility Index 2026
Semrush has launched an AI Visibility Index. It tracks which brands ChatGPT, Gemini, Google AI Mode and AI Overviews mention, and which websites they cite as sources. It’s built on more than 180 million AI answers across 22 industries and 9 markets, and updates monthly. Always worth a gander.
Meta’s Threads Now Has More Monthly Users Than X
Worth taking another look at Threads? On Meta’s earnings call on 21 October, Mark Zuckerberg said Threads has overtaken X in monthly active users. He didn’t give a number, and Meta’s last published figure was 500 million in June. Omor Ibne Ehsan at 24/7 Wall St. points out that X still leads on daily web users (141.7 million against 7.6 million for Threads) and among US mobile users, which is where advertisers pay most. Meanwhile, time spent per user on Threads grew more than 20% year on year.
PS: Threads exposure in Google Discover is MASSIVE btw.
Norwegian media companies take Meta to court over AI scraping terms
Charlotte Tobitt reports that Norway’s publishers, including Schibsted, NRK and TV2, have asked an Oslo court to block Meta’s new terms. These would give Meta the right to scrape and reuse content from any publisher that uses Facebook or Instagram, including paywalled content and content that never appeared on Meta’s platforms. Meta has pushed the start date from 30 October to 20 November after the backlash. Danish and Swedish publishers have gone to their competition authorities, the CMA is in urgent talks with Meta, and one local paper in Røros is simply leaving Facebook. Meta says it will respect robots.txt blocks on AI scraping, but that isn’t written into the terms.
Instagram Tests Videos up to 90 Minutes Long but Unlike YouTube It’s Not Sharing Ad Revenue
Todd Spangler reports that Instagram is testing horizontal videos of up to 90 minutes with a small group of US creators, and has launched on Apple TV. Adam Mosseri admits he was against long-form video for a long time, because it pulls Instagram away from friends. What Instagram isn’t doing is sharing ad revenue the way YouTube does. It’s funding some creators directly and “exploring” the rest.
Did you know that 50% of app searches actually start on Google, not in the app stores?No, I did not know that, but Cindy Krum is going to tell us all about it.
The Real Kevin Bacon
I love The Pudding, but I also find this hilarious because I think my wife is actually in love with the man 😍. Publishers sitting on their own datasets could do far more of this. I remember the days of Martin Belam’s UsVsTh3m - why aren’t news publishers doing more data-led experiences like this on a massive scale?
Until next time,
Steve