On this week’s episode, we hear from Sarah Ebner, Executive Editor and Head of Newsletters at the Financial Times. She tells us about her role leading the newsletter team at the FT, and the value of newsletters in subscriber acquisition and retention but also as paid products in their own right.
In the news roundup we take a thorough look at what the integration of ChatGPT and Bard into search results means for news and magazine publishers. In the news in brief, the Mastodon Bump has levelled out, DC Thomson announces 300 job cuts, and we ask why subscription revenue is outperforming expectations.
Welcome to your first episode of Med.A.I Voices.
Here are some highlights:
The newsletter set-up at the FT
We have 35 newsletters at the FT, 16 of them are Premium. Newsletters are really important at the FT, and I hope have become even more so since I’ve arrived. We use them to acquire, engage and retain subscribers. I think anyone in the media business now knows how important newsletters are for anyone who’s got a subscription or membership business.
[The Head of Newsletters] is a real combination job, which is one of the things that’s appealing about it. It’s still got journalism in it, but there’s an awful lot of other things that you can learn a lot from working with other people across the business, which you don’t necessarily do in many other jobs in journalism these days.
When I came, I think newsletters were seen more as a very good editorial product, where the Head of Newsletters was doing a curation job. Whereas I see it as you having to do loads more to make them work. You can have great editorial content, but if you don’t know how to find or sign up for the newsletters, it’s not really worth it. So you have to work with people to make [newsletters] easier to find; big discoverability things, redesigns so they’re easier to read, that sort of thing. There’s always more to do, but I think we work pretty well.
Creating two-way communication
We know that readers really value [our newsletters] because we get a lot of feedback – we get loads of emails back from them. One of the great things about a newsletter is that direct communication with a writer, and we always make sure that you can contact them back in our emails, either by putting an email address in or having a reply-to email address.
So we get a lot of responses from people, and they love that personal interaction. They love contacting [Associate Editor] Stephen Bush and saying, “Oh, you’ve said you’re coming to Edinburgh, I’d recommend you go to this restaurant,” and he replies!
But there’s also fantastic bespoke original analysis and content. I’d say that’s a real strength of the FT, and in our premium newsletters – which are truly premium – you get Robert Armstrong telling you about the markets every day with his own spin and his own analysis, which is amazing. But so is Martin Sandbu in his weekly economics newsletter, which is so high-level and so valuable to people, we get emails from very important people saying “This is great, this is really important.” You can’t get this stuff unless you sign up for the newsletter, so there’s a definite value proposition.
Different strategies for different newsletters
Different newsletters can have different strategies – and I think should have different strategies. The idea that there’s one that works for them all is ridiculous, because they will have different aims. Our premium and standard tend to have different aims: whether they want to send you back to the site or not.
The added value is so important. And there’s a sense of utility – what’s the utility in the newsletter? In some newsletters it might be, what tips are in there? What can I practically do with this information?
In fact, we’ve just launched our first newsletter course, MBA 101. It’s the first time we’ve done a course. It’s six weeks, and it tells you how to get into business school. So it’s genuinely useful, it’s evergreen, so you can sign up for it whenever you want. I’m excited about that because it’s a genuinely useful thing, and should get to different audiences that wouldn’t necessarily come to the FT. We have really good MBA rankings on our site, so this seemed the perfect thing to tie in.
Handling Apple’s privacy changes
We originally artificially deflated our open rates – in that we counted them as zero – which was problematic as it meant that growth rate went down a lot. We’ve changed that now, and we’re one of the few people who have verifiable open rates. I like to be honest about things. We count Apple users in our sends, but we don’t count them on the open rate. So our open rate is based on Android.
We know from before the Apple changes that our Apple users were actually slightly more engaged, so we’re probably deflating by a tiny bit. I think people use this and they go, “It’s fine because we compare all our newsletters to each other.” But actually within the industry, I’m always suspicious now when people say “We’ve got this amazing open rate,” unless it’s verifiable in some way.
Top story: AI and the future of search
This week, Microsoft announced that the new Bing would have ChatGPT built in.
- Microsoft has invested $10 billion into OpenAI, the start-up that created ChatGPT
- Bing cites its sources and links to them in a “learn more” section at the end of its answers. Every result will also include a feedback option
- For some queries, including those about shopping, the chat feature will show ads
Google’s demo of Bard – its own version of ChatGPT – cost it dear this week when it ran an ad demonstrating the tech, and Bard responded with a factual error. The mistake (the James Webb Telescope didn’t discover exoplanets…!) caused the company to lose $100 billion in market value on Wednesday.
- Critics argue the fact could have easily been checked with a quick…uh…Google.
- This goes back to what we were discussing a few weeks ago with CNET using AI to generate articles. These mistakes are creeping in when the technology and its results are at the height of scrutiny. We’ll almost certainly see bigger and more critical misinformation coming out as human oversight eases off.
For the publishing industry, this poses some difficult questions. Nieman Lab asked one this week: how will AI-powered search impact publisher revenue?
- Author Joshua Benton thinks this is the most important shift in consumer technology since the iPhone. “Will these new interfaces simply mean a modest improvement in 5% or 10% of our search requests?,” he asks. “Or will AI become the dominant way we interact with information online?”
- The more questions Google and Bing answer without a click, the less traffic and therefore ad revenue those ‘answering’ sites will get. It’s a big shift from link-based search to algorithmically-based search.
- Fears for the future of publishers in search were expressed on LinkedIn by FIPP-congress organiser Di5rupt’s Cobus Heyl. He wondered what it would mean for publisher attribution if search results are ultimately generated by AI from an aggregated dataset.
- How do you even begin to credit providers and provide links back to original sources when everything is a source, and it is all condensed into a few paragraphs?
- Earlier this week Chris wrote an article for The Drum asking if advertisers should get discounts on AI-generated articles – because in what sense can they be considered ‘premium’ at this point? The received wisdom currently is that, provided the articles deliver the same value to audiences as a human-written piece (and that’s a big if) then they should be priced the same.
News in brief:
- Bad news for anyone looking to Mastodon to act as a direct replacement for Twitter. With too much chaos at Musk’s Twitter to discuss in any detail here, many people are understandably looking to other platforms to fill the void. But, as this article on Wired makes plain, the early diaspora of Twitter users to Mastodon has not continued at the pace of its early days, suggesting that the service is not a direct replacement.
- DC Thomson employees were told at an all-staff meeting that the company needs to plug a £10 million gap amid moves to “reshape” its portfolio. This meant 300 redundancies and 40 magazine closures. We think of DCT as a Scottish publisher, but 150 of those redundancies and up to 30 magazine closures will be in Colchester with the complete closure of Aceville, acquired in 2018.
- According to FIPP’s latest Global Digital Subscription Snapshot, publishers across the world are growing subscriptions despite economic challenges. The NYT Group has seen 11.8% growth across the quarter, The Pioneer 88.2% and Substack 50% across its newsletters. Local media is also resisting pressures, with subscriptions continuing to rise across newspapers of all sizes.