3 ways AI will shake up your reader revenue

Welcome to the Paywall Podcast, where we discuss paywalls and paywall strategies for news and magazine publishers. This episode of the Paywall Podcast is brought to you by Leaky Paywall. Leaky Paywall is the most flexible WordPress subscription platform. Find out more@leakypaywall.com and now for today's episode.

Today we're going to talk about three ways that AI will shake up your reader revenue and how to take advantage of it. Welcome, Tyler. Welcome back to the Paywall Podcast. Nice to have you here.

Hello.

Hello. Ready to talk about artificial intelligence chat GPT and how it will affect.

Let's do it.

All right, let's do it. So a big part of the news, and I'm going to just jump into what's going on in the news here is Time magazine. They announced that they were going to stop paid subscriptions. And I'm kind of picking on Time a little bit as a. What you should not do in the publishing world as far as publishing news. And, hey, we were a longtime print subscriber of Time magazine. Fantastic publication, loved reading it. But the digital world unfortunately doesn't fit with the model that Time sort of brought from print to digital. They decided to go broad with their news stories and going broad with your news in today's Internet, where it's all about the niche is a really a rough road. So they drop their paywall, they're going to lean into advertising. Not sure what the heck is going on there because ad revenues are not going up for digital publishers. They are going the opposite way in general as a trend. So we could talk about Time and what they should have done, but it sort of segues into my first point about, you know, well, let's talk a little bit about AI. Well, what, what, what is AI and what does it do? It's a way of automating lots of simple and somewhat complicated things that you would do as a publisher. Content being, you know, creation of content is one of them. And that's probably the biggest. I don't know what you see, Tyler, but I think that's sort of the biggest application right now is you can type something into ChatGPT and say, hey, write me a story about this, or write an article about that. And it will just do it and it'll produce an okay article and the grammar will be good and that's fine, but it's not inspiring, it's not personal. It's. It's a bot. And you can get really good at it and feed it prompts that. That let you output better and better content. But the problem with, with AI and publishing right now is that if you're in a, if you're in a broad segment and you're producing lots of sort of gen, more general, broad articles, AI is going to eat your lunch. Your, your really your, your best bet as far as producing content is to go the opposite direction that time is going. And that's niche down. That's create content. Content that is super tight and relevant around a very narrow topic area. So what's a very narrow topic area? How about local news? I'd say that's a very narrow topical area. You're covering what's happening in your town, in your area. That's it. Nothing else. Maybe you're the best at it. Hopefully you're the best at it. How does AI impact? Let's say I start up an AI news publishing business and I'm going to go tackle Morganton, West Virginia, and you were a publisher, Tyler, what would you do?

Well, so I've got a publisher. Let me just start with this. I've got a publisher, I believe, in the suburbs of Chicago, who's actually using AI as a tool for generating content. They don't use it a lot, but it's interesting. You're talking about the broad content area. That's where they're using it. That's right, precisely. It's helping them do research on these broader topics that aren't necessarily their wheelhouse. And so I think one story in particular, in particular was about gun control or something, and they were talking about the broader issues around guns in America and all that. And they were using AI to help them do some of the background reporting on it. And then they were localizing, you know, that. The context around that. So that's kind of where I see, as far as local news goes, using it for background and research. And then obviously you're still going to have to have to do the local news reporting on the ground and you're still going to have to fact check all the stuff that those AI bots throw out at you. So it's not like you can just throw in some prompts and assume that everything in it is true. It might sound great and look great, but AI is only as good as the data that we have out there. And so it's, you know, proceed with caution. But I don't think it's the end of the world.

Well, it's, I mean, it's the dawn of the computer has always been garbage in, garbage out. Right. So the human it's almost like a magnifying glass. Right. Like, sounds like they're using it for, let's say, an article on gun control. And that's a very, that's a focused article. It's kind of broad, but it's, it's a focused topic. And then using AI to do the research behind that article, I think that's, that's a great tool. That's, that's a time saving tool. Any tool that saves you time today is, is a win.

Yeah. And they're very explicit too about this article was generated with the help of, or with the support of, you know, Chat GPT or whatever labor of AI that they're using. So.

Right.

I think that's also an important part of it. If you do plan to add AI to your workflow, and especially as a local news publisher or even a regional or national publisher, you should probably be letting people know that this content is generated by AI. But you know, you're also fact checking it and making sure that it's accurate because obviously at the end of the day that's, that's what the job is as a journalist.

So be transparent. For sure. Yeah. You know, it's, it's a tool. I got into a discussion with Colleen, my wife, and we were talking about, I had written an article with the help of AI as a kind of a parody on publishing and she really liked it. And then, then I told her I used AI2 and I was very specific about the prompts that that was going to build up the skeleton of the story. So I basically wrote the skeleton of the story and then had the AI create the story. It's almost like hiring a writer. And I pay for chat GPT4 and, and it works really, really well. She was pretty horrified about. So, you know, if I had told her up front, hey AI, you know, I had, I wrote a story with AI, that reaction part would have been a little different. Yeah, you would have known coming in that. That that was that way. Yeah. Yeah. So, so I'm just getting back to. So if you're in a niche and you're, and you're niching down on your, on your content and could be, could be local news or could be. Maybe, maybe you write about control, gun control. Maybe that's your niche. Maybe you write about fishing in, you know, Massachusetts. That's your niche. Use the use AI. Yes, you can create it, you can use it for content creation, but you got to be careful because it does, it's not as good as what you can write yourself. But the research side, I think that's a great, that's a great point. Yeah, yeah, yeah. Okay, so niche down. Stick to your niche. You know, we think about, especially local news. Like you absolutely need to focus on your local area like that. It's, it's almost, you know, some of the most successful news publishers that, that have, you know, local or regional newspapers. You know, the rule is if it's, if it's out of bounds, if it's out of your town, you know, if a nuclear bomb goes off next door, you don't write about it. I mean, I know it's a little extreme, but you write about what's going on in your town. You mentioned people in your town. You draw your readers into it and that segues into that. Then my next point, the second point, so niching down, yes, you have your expertise, you build your walls around that. But now you also get to build a relationship and strengthen your, the relationship with your audience. And I think this is going to hurt magazine publishers a lot that aren't, that are still really focused on advertising and haven't embraced reader revenue yet, which is a big chunk of magazine publishers. But, okay, you can crank out a lot of content using AI and it's maybe not going to be great. Maybe if you're really good at it, it's going to be great. And that's, that's okay too, as long as you're upfront about it. But that relationship, and this is, again, this is something that every part of publisher knows. Yes, we want to, we need to please the readers to some level. I think news publishers really lean in on that reader relationship, and magazine publishers really don't lean in on the reader relationship for the most part. Some do. The ones we work with do, and it's paying off big time. But many are really focused on leaning into the relationship with their advertisers and at the peril of ignoring their readers. And I'm not saying they completely ignore their readers, but advertisers are driving the train. Then you got to be really careful because those readers are going to evaporate if you don't really, really take care of them. And so the relationship between the publisher and the reader, I think becomes even more important than it already is, and it's super important now. But with AI, you have a brand, right? You have, you know, you have ways to connect directly with your reader, the newsletter being the big one. And of course you have your, you know, your mission and vision and, and your focus and all that. But I think Leaning into that relationship, making sure your newsletter, you know, is personal. Really? Yeah, Your business, but it's personal. We connect with humans. Making sure you're, you're, you're, you're building that list with a free registration, which we talk about a lot. Super important. And that will help create a moat around your publication. The bigger your audience. And this is where everyone agrees, magazine news publishers, niche publishers, everyone wants a bigger audience. That is like a universal truth. And so working on capturing those casual readers and growing that list and putting you in control of communicating directly with them. And then once you get that straightened out, you look at, well, some, I don't know, Tyler, if you see publishers that are leaning into commenting, you know, like if you're running subscriptions, that's a great opportunity to let subscribers, you know, talk about your articles, comment, engage with each other for better or for worse. But if you're a paid, if you're, if you've delivered a credit card and you're given the ability to comment, most, most of most of those subscribers tend to play nicely from what I've seen so far.

Oh, for sure. Especially if the only way that you can comment is whether or not you paid or not. I mean that's a great screener and obviously your name is attached to the comment, so.

Yep.

Yeah, that's a great. I've got a couple of publishers who are running comments and for the most part it's, you know, it's not a big deal. I mean there is the occasional blow up, but for the most part it's. Right, it's totally fine.

Yeah, it's a great engagement tool. Social engagement is a motivator for paying subscriptions. And if you're, if you have delivered your information and a credit card number, you tend to not be a total jerk. Yeah, yeah.

And one thing I wanted to add, this is going to sound probably super simple and maybe, maybe obvious, but as far as like personalizing the experience, one thing that I do with all of my publishers from the very beginning when they come on board is making sure that they have photos, avatars of all of their authors that show up in the stories. So like I want to see, you know, Jane Smith and John Doe, like I want to see their photo and the local news story. So just a small little thing that you could do in your spare time, personalizing the overall experience. And it helps. I mean the publishers that do it and have nice photos, I mean it helps.

You know, I've seen, I've seen, we've worked with larger publishers that don't have their act together on staff photos. And it's an effort, I get it, it's an effort, but it's, it's really important. And the ones that are really brilliant at it inject some fun into it. Like I remember I politics. Like if you go to their site they have like the serious like staff photo and then if you rolled mouse over the photo it turns into their, you know, super goofy, you know, like super personal. Like I, I love cats so I got a cat or I right got my, I got my motorcycle helmet on because I like to ride bikes or whatever it is, you know, makes it more personal. But that's part of the engagement. It's really important. Super important. New tools like Subtext where you can send SMS messages to your readers directly and they can SMS you back and you can. And manage that through a dashboard adds personalization and actually adds a whole additional revenue channel for you to take it that far and that, you know, that engagement with email messaging, you know, being personal commenting, I don't know if I forgot anything, but those things will build your moat and keep people, you know, attract and keep people with you as a publisher, regardless of how crazy AI gets. And I'd say, I'd say the time is now. You know, we sort of leads me into, into one of my next points is we get comments from publishers about like search, like dropping off. Like, oh, AI is going to take over Google search or, or you know, when you're going to get like you're gonna, we're all using AI to search for stuff. That's not really true. Yeah, maybe a little bit. Mostly AI is like information generation and content creation and like you said, research. And the, the reality is, is like back to time, like time is becoming more and more important as, as things accelerate. And we still, you know, like when we approach the website, we're scanners, right? Like we're, we don't really read content. We're not ready for big blocks of content. When we're, when we're searching for something, we want quick bullets, we want quick links. And that's why Google, you know, search works so well because you get these quick links and you get a quick excerpt summary and you can scan down the page. If you get to the bottom of page one, you might go to page two. You probably won't go much further than page two. And, but that's, that's, that's the world we live in. What's AI going to do to change that? Are they going to like produce the answer, the one answer. For us to read and ignore all the other answers. No, nobody wants that. We need trust. We need to, we need to see sources of information.

Right.

We need to, most importantly, we need to scan the data. So SEO is going to be around for a while, maybe in some form forever. AI version of SEO, which Google is. Yeah. Now releasing. But. Yeah, so that's one, one thing I wouldn't get too upset about at this point.

Well, as a local news publisher, I mean, you are the source of local news in your community. So as far as like AI taking that content and rehashing it back to someone, I mean, it's going to have to get around your paywall. There's a lot of barriers in front of AI. So I don't, I think going back to what we started with, just using it as a broad tool for research and, you know, obviously making sure that that research is accurate. And you're still going to have to do the legwork of city council meetings and all the things that come with being a local journalist. Afraid to say.

Right. But you know, there's a good point. Let's say you, let's say you do city council meetings, right. Which is, which is great for local news and people want to pay for it. Right. You can create a framework where you use AI to help produce the content. Right. You know, I don't know what the legality is on recording stuff and transcribing stuff, but you could use AI to build summaries off of, off of interviews, for sure and things like that. And like, let's say sports scores. Right. Like another thing, like if you find, if there are pieces of content that are. You're writing over and over again, it's repeatable approach. That's a great application for AI where you can produce content much more efficiently. You're still doing all the legwork, but AI is sort of helping you speed along the content creation that you're inputting it. So good input, good output. Right. Yeah, I'm just checking my notes here. So we talked a little bit about search. Oh and you know, there's always the nuclear option for, you know, will, Will AI steal all my content? That's, you know, that, that fear and that's you can lock content down. Now if AI steals content, which is, which is starting to happen. I read about a publisher where that was a very, very niche publisher who was also found out that there was a website that had essentially stolen his content and was rewriting the content. You can always lock down your content. That the nuclear option where you, you start, you go back to the old model of like, okay, these are premium articles. We'll lock these down, or we'll just lock everything down. But of course, you're killing your ability to show up in search and you're killing social sharing. It's not the option I would recommend.

Yeah, it goes back to, like, when publishers worry about whether people are going to share logins with each other or whether they're going to clear their cookies and get around the paywall or they're going to change their IP address and get around the paywall. Like, some of these things you just like, have to accept as a, as a thing that will happen, but it just won't be the, it won't be the majority of people. Most people don't have time to mess around with that, right?

No, they don't. You know, and you have, like, as a publisher, you have a brand, you have an email list, you have your social channels and you have engagement. Right. You have the trust. So somebody just going to go to another site to read rehash stuff. Yeah, maybe there'll be a few. But if you really want the source, like you mentioned earlier, you become the source of content then. And you, and your, your was reading an article this morning about like, what, why do people pay? Right. And I shared that with you and it's like, well, it's because you have a great mission. You're in, you know, independent journalism. And we, and here's the pricing, here's the simple pricing. And if we don't get your support, we won't survive. Like, we need your revenue to survive like that, that's messaging that resonates. That's the, you know, it's part of the engagement of the reader. Right. And we do want to support independent journalism. It's very, the data was very clear in the report. People want to support independent journalism. Yeah. All right, I got one last thing I just wanted to share before you wrap up about AI. And I think if you haven't been using AI in your publication, now's a really good time to start playing with it. It's, it's, it's fun once you start, you know, just sign up, go to chat, just Google chat, GPT sign up. It's free. Sign up for a free account. You might want to pay the 20 bucks a month and pay for the pro version, which gives you chat GPT4, which is quite good. And we'll let you do what I'm going to, what I'm going to Mention next, but it'll give you a feel for how it all works and what, how you, you could put it, put it in action for your publication. So one of the things that, that we use it for, and this podcast is a great example of it, is we use it to build articles on our blog. So here we, here we're in the middle of the podcast. We're recording the podcast video version and the audio version. And the workflow for us is we take the, take the video and put it on YouTube, that's fine. And then we, then the podcast goes up on all the podcast channels, but then we throw the audio into a tool called descript, which translates the text. Doesn't translate the text. It produces the transcription of all the text. I think it can, it might actually be to translate, I'm not sure.

Anyway, there's still all this jargon that we talk about.

Yeah. Anyway, it's called the script. Does a good job of transcription. And of course, a raw transcription of an interview like this or, you know, that you might do, or a podcast is pretty rough. Like, it's hard to read. You can't really get through it. But we take a copy of that text, we throw it into ChatGPT4 and we ask it to write a summary of, of what we talked about in this. And it works great. A nice summary comes out. Yeah. Gets a little editing, sure. Always, always need to edit. And then it gets posted as an article along with the video, along with the podcast, the audio. So if you're doing a lot of interviews, if you're doing video production in your publication, you definitely need to get Chat GPT4 because you can, you can post long transcriptions into Chat GPT4 longer than Chat GTP3 and then use, you know, use it some transcription software of your choice to create those. And, and you will have nice summaries that you can publish along with your videos. And why do I say all this? Well, remember we talked about scanning. Right. We, we are, we get to an article and we realize that when people come and look at one of our articles on our blog, you have lots of different people. Some will, hey, I'll watch the video. Some let me check out the podcast because I'm a. I love podcasts. And some will simply need to scan and probably a big chunk of it, of visitors need to just scan some text to decide whether they're going to listen to the podcast or watch the videos. So that little summary, it's a short paragraph, lets them quickly read about what this podcast is about and then decide to dig in deeper, maybe subscribe for future podcasts. It's a critical element in encouraging your readership to watch that video or listen to that interview or whatever it is. And the workflow is very quick when you use ChatGPT for it. Okay, I think I'm done. Anything else we want to. Want to add to AI?

I think that's it. Should we tell everyone that you and I both are AI at this moment, or should we leave that alone?

That's pretty funny. So we are not. This is really us, however, is it starves for the rest of us. Rob over there, he did a bit. He did a video on where he. There is a service that. That will watch a few clips of you talking, and it will. It will then output a lot of you talking. And it's. It's a little eerie how it's not perfect. Like, you can tell it's a bot, but it's pretty good. Yeah. Yeah. It's just. It's a little. So that kind of weirdness is coming. Yeah. Anyway. All right, Tyler, onward. Embrace AI, play around with it and apply it to your publication. It will help you catch you next time.

Thanks for listening to the Paywall podcast. If you'd like to get in touch to discuss subscription Strategy, go to LeakyPaywall.com see you next time.