The Newsroom Is Gone. AI Filled the Vacancy with Linda Zebian of Muck Rack

If someone looked you up on Google, they went to the internet and they typed your name in the search engine, what would come up? Would it be Google? Would it be social media? Would it be something driven by AI? Or would it be a Reddit thread? This is something that I told my guests just now that is happening in real time in my work in crisis management. From a call that I had this morning, and that's why I'm excited to talk to Linda today, because I want her to answer a question for me. Linda, welcome to the podcast.

Well, thank you for having me. I'm so excited to be here, Molly.

So Linda, you are the VP of Communications for Muck Rack, and I didn't realize this until I looked at your bio. We've been working together for a while. You were at the New York Times.

Yes, I was there for 10 years in corporate comms.

So you are someone who can speak to the changing landscape of media quite well.

Oh yeah, I was there for like the whole BuzzFeed, HuffPo vibes that, you know, have, have totally shifted now. But you know, like those digital native startups that were threatening traditional news. So I've seen it all.

Where have we landed right now? We're June 2026. Where are we in that media landscape? Who has power?

Yeah. More than ever, nobody has, nobody has the power and it's really fragmented, right? You see, you hear that fragmented media landscape. Um, it's just the kind of a buzzy kind of obnoxious thing that PR and media people say all the time, but it is true. It is very fragmented. People are going to verticals and experts, subject matter experts and influencers, if you will, who share their views, who have their same beliefs as them. And so they kind of gravitate towards these very specific Substackers or podcasters or YouTubers, columnists, et cetera. And so it's really, really starting to be very, it's very targeted. It's very niche. And I think what that means for the PR and communicators of the world, it's not necessarily a bad thing. I actually think we have more cards to play, more avenues to take with our stories. We just have to be able to, you know, identify the best places for that. And what best is now is also changing.

Well, that was the question I was going to ask you. Is this good for PR?

It's certainly more difficult to place stories than it, than it's ever been. And I think that just is going to continue to increase because what you would call like a, a trained journalist, that number of folks is decreasing. The newsrooms are shrinking. Local news is in absolute peril. If you saw McCracken Rebuild Local News put out a report just this week that We have 80% fewer journalists, local journalists, than we did in 20— in, uh, 2002. That alone, the ratio of PR people to journalists is, is, is definitely shrinking. However, as a PR person, there are other ways, there are other avenues, there are other kinds of content creators or newsfluencers or whatever you want to call it that are certainly part of the conversation that we can, we can work with, um, to get the stories out and, or to combat stories as well.

That's a very interesting statistic of what you're saying about the number of reporters that we have access to, particularly, particularly at a local level. But certainly a lot of reporters, as you mentioned, are still ending up online. Has Muck Rack kept up with that technology?

Oh yeah, absolutely. I mean, Muck Rack doesn't just have traditional, what we would think of as traditional media folks in our platform. We have all the Substackers. We just increased our podcast monitoring significantly. We have news producers, you know, television producers and writers, and we have all, all the guys, all the newsfluencers, bloggers, they're all there. To just have traditional news would be not serving the PR and communication audience, which is like, that's our—

This related to a client of mine today and our conversation in the back of our head, because I think I was thinking What I'm telling this client right now, I wish I had Linda sitting on my shoulder right now with this letter because they were dealing, uh, certainly a crisis, professional one, but it bled into the personal space because it got viral and messy. People mobilizing online and some, uh, kind of those social media vigilantes that are for causes, but can cause destruction. So some people genuinely are mobilizing to do good and to do better. But I find in my work, some people want to opt in to optimize the friction out there and monetize, or they have accounts and things like that. So I was talking to this client because this was such a mess. They ended up losing their position and it just So the discussion went towards what the reputation online, and this was a company with a big presence, we'll say online presence, but the client who I was, that I was speaking to also had a media presence, a lot of traditional media hits out there. So the conversation drifted. I don't even want to say it. I abruptly moved it towards this. And that is what people see when they Google her name and what comes up. So we talked about SEO and that's search engine optimization when you go to Google and that's kind of school now, correct? And I mentioned G-E-O. So Linda, give me your definition of G-E-O and how it differs from SEO.

I think the most simple way to think about G-E-O and SEO is that SEO was essentially a ranking on a list. And you had a lot of control over SEO with your owned content, right? It was owned by marketing, your own site, right? Who you are was based on what your website looked like. GEO is all about being in the answer. These synthesized AI LLM answers. When we're talking about LLMs, we're talking about Gemini, ChatGPT, and Claude primarily. So it's how do I get in the answer and how do I earn trust in the answer? Those citations and sources come from a vast variety depending on what LLM people are searching on, but it's about controlling that answer.

And you brought up a word that was a word that came up in conversation with a client, and that was authority. And because I'll just say we're on the cuspy Gen X boomer, and so I'm, you know, explaining the difference, you know, between the two. But the concern there was, well, there are a group of people who are out to get me, and if this is announced on this platform, they're going to screenshot it and then they're going to hear it. So the conversation then went on authority. And I, as I was speaking about GPT, what I'm noticing now in the LLMs is when you do a name or a company, authority comes to the top. Still source Reddit, they'll still source Facebook, they'll source social media, but it always comes with a caveat and say, we can't prove this, or this may not be credible. So tell me, why would LinkedIn or New York Times land at the top search as opposed to a Reddit?

We had this question. We were building a product for AI visibility for PR people, marketers. And we were like, well, how do we know what AI actually reads? How is AI— how are the AIs sourcing? And so as we were building this product, our data engineers decided to do the research. And then when they were doing the research, we said we should publish this. So that's kind of how the What Is AI Reading, um, kind of mania last summer happened. And we've put out this report 3 times in a year. It's called What Is AI Reading? You can get it on the Generative Pulse site if people are interested in it. But what that research found was that 99% of what AI engines are citing— this is across 25 million citations, I would say the biggest study of its kind— come from non-paid sources, owned media and earned media sources. This is the difference between like traditional marketing.

You mean by both for people who aren't in the industry, Sinead?

Sure. Paid media, we essentially mean advertising, advertorials, sponsored content, these kinds of things. Okay. And the rest of it, which we would call non-paid media, includes journalism, corporate blogs, like, you know, Molly McPherson dot— well, that would be more of a journalistic, but like, you know, muckrack.com is a corporate blog. Target's blog or whatever, um, press releases, academic research, government sites, social, encyclopedic, that kind of stuff. That's the rest of the, um, non-paid media bucket.

And here is the argument, even that is folded into what you just said. At the beginning, at the top, we talked about how the media landscape has changed so much and how it's very difficult to reach, uh, journalists now. Yet you just cite press releases as being in that list of authority. So you have just cited that there is still a re— why we want to send out press releases, correct?

So if you look at the pie chart of this report, I mean, I think the people really are focusing a lot on like things that you mentioned, Reddit, LinkedIn, and Facebook and other, these other sites. But if you really, I mean, if you really go down to individual sites, like For example, social, which includes Reddit, is only 2.9% of citations. Okay. So that means that Reddit, along with LinkedIn, Facebook, TikTok, YouTube, all of them, like you could figure Reddit's going to be way less than 1%. Okay. So when people are overly concerned about Reddit, we have to kind of step back and be like, yes, it, you know, it does, it is cited, but, but the models want, want third-party credible information. Okay. And to your point about press releases, the press release accounts for about a little over 1% of the total pie. It's not insignificant. It is somewhat significant because that's a 1, like it counts. I would say if you wanted to compare them as much as Reddit, right? Because if Reddit's less than 1% and press releases are 1.1% of the whole pie, you could see them as like LinkedIn press releases, you know, YouTube press releases. Okay. So let's think about it that way. The big chunks we want to think about are journalism, corporate blogs and content, the aggregators and the encyclopedias, right? These are the big chunk. And then owned media, which would be your own site. Those are the big chunks that we want to think about. But to the point about press releases, press releases are going to be cited. You want— and you— they're still valuable, right? The LLMs pull from them. So to get them to be citable, you want No showy, flashy nonsense, fluffy words. You want purely facts, pure statements made, you know, quotes and direct statements made by high authority people without like the fluffy nonsense. You want to put a nice summary, an AI summary at the top with bullet points so they can really just pull right out of that press release. Short, punchy, why this matters, when is it happening, dateline, all of those structural things about a press release, the LLM loves. Because it feels like very authoritative and very—

let me ask you this. So for years we've written the same format press release. I mean, my goodness, we go back in college. It's the same one that it is now, but when you write it, put that AI summary at the top, the formatting of a press release, a traditional press release change that can improve AI picking that up in an LLM or an LLM picking that up. And what is it?

Yes. So what you want to do is beneath your subhead of your press release, you want to put the word summary or AI summary at the top and then bullet out with your data points, no quotes or anything, the summary of your press release in short punchy bullets so that the LLMs can pick it up. Same with blogs. I mean, you want to do the same thing with your blog. You want to do the same thing with your own site.

Now that is fascinating. So we're all showing that, you know, earned media is what shows up in search engines, but it also shows up in the LLMs and the AI. And this fits what I was calling someone who is crumbling, eh? Uh, because all they can see is Reddit and it's very hard to get through with them. But if you show them the numbers, like you said in the study, It really does matter because if you go to a ChatGPT or a Gemini, you will see this type of summary. I'm noticing as someone on AI and using it for reputation hacking for years now, I noticed that social media had so much more weight a few years ago. Not only is it lesser, but it also has like an X next to it. Not proven. We could corroborate this, this, and it's not necessarily going directly to— it's just giving like a small summary.

Exactly. And I think that, you know, of course, if we go to Reddit and we go to ourselves or we go to our company and we try to see what people are talking about during a crisis, and you're— that's all you're going to see. But like, if you take a step back for a minute and realize that like most people are not out there Googling you or Redditing you or searching for you or your company. But if you go look for it on one of these sites, yeah, you're gonna see that in the moment. But like over time, you know, these synthesized answers, you're gonna just dissipate, dissipate, dissipate. The way to counter that little sliver of Reddit is to use the other categories and the other outlets, media, your own site to, um, counter mis— mis— disinformation, put out positive stories about you and your brand. Surface these other kinds of things so that, because those are the majority of the conver— where the conversation is happening. It's not on Reddit, but if Reddit has something negative, you can certainly kind of, um, balance it out with positive, uh, in earned media and other, and other ways.

Absolutely. And so what you're saying here is so important for communicators right now and for leaders to understand. Everybody put all their energy into social media. I think social media is what drives culture and what drives news, drives conversation. It's incredibly powerful. If you look at the reputation management aspect to it, we cannot overlook our websites because of the GEO and the authority, correct?

Absolutely. Our websites are owned, but also are earned because these AIs, which is where everyone's going to search now, even if you go to Google, you're hitting that AI mode or whatever it's called, AI. Answer because you want the best answer and it comes through AI. The best answer does not come in a listed, you know, sponsored nonsense Google search. So what those AIs want is consistent, high authority records, right? Records of whatever is happening. And that happens through earned and through owned. And they don't want to pull from Reddit, but if Reddit's the only thing that's kind of wah wah wah about you, they're going to pull that. But if you've got something on the other side of it, you've got more stories to tell that are the actual messages you want to convey, it will pull from a more authoritative, more credible site, like a news site or like your own site.

Speaking of Reddit, I went to Reddit and I went to some of the threads about GEO and asked a question. I thought it was such a good question. So I have to, I have to ask this. They wanted to know which had more authority on the own site on a website. An article, so a blog, or FAQs?

From our research, it depends on the query. So if somebody is asking about a trend, what's the cool new running shoe in 2026, for summer 2026, that's going to cite, um, journalism primarily because of its trend-level authority, but FAQs are gonna come surface for more direct factual queries, right? Does Nike make a court shoe for summer basketball? What's gonna come up is Nike's website is gonna come up. So fact-based queries, you're gonna have owned content coming through, and FAQs are great for that. If you're asking about trends, a trends piece or something like, you know, more fleeting than that, it's gonna be a blog or a news site or something like that.

And I'm looking at it through my lens in crisis management, working with clients. We need to decide where are we responding? Something is dusting up and it's, it's messy, like on social media, and things are starting to go viral. When I'm working with a client, they want to know what we do. Do we ignore it? Do we post something? Do we say something? Where do we do it? I'm constantly scanning what is the best. And I have to be honest, I've changed. I have changed completely in how I, and how I do things. You know, sometimes you would play nice, Meta, if, if someone was saying something on Facebook and it was wrong, you wanted to correct misinformation, we were correcting it on our own Facebook page. But now I'm, let's get out of that algorithm altogether and let's find GEO and let's go to our own, you know, our own places, which is our website. So now I was thinking I would typically then do FAQs because the FAQs, I would take what everybody was complaining about or misinformation, turn it into a question so we could get the answer. But then I was thinking of GEO again. I'm wondering, are they going to pull the FAQ section or will they pull the article? And my gut was telling me, I think we have to put these articles and you verified that with the study. When we're dealing with product, when people have questions and queries, there's a value in FAQs, right? Like you think of Gemini, boom, boom, boom, it comes up right away. But in the reputation management piece of it, for anyone listening or watching, when they're thinking about, oh, well, how do we respond to something? It sounds like what you're saying is media statement and news article and embed in either one, correct?

Absolutely. I think the lesson here is everywhere. Everything everywhere. Whatever you've got that supports your narrative, you post it and you post it with words. You don't post just a video. You post a video of your CEO addressing the company about the issue with the transcript. Always. They can't parse a video. So you need all your transcripts. You need all your words. Don't put a PDF up there. No, no, no, no PDFs, no PDFs, no graphics, none of that. They're not, they're not going to parse it. I think that's the beauty of these like pages and pages of copy that nobody really ever sees when they go to your website, but the LLMs are scraping. So if you, for example, if you want to become an authority about a certain topic as an organization, you just put that long form stuff right on your website. No one's actually going to read it, but the LLMs will synthesize it.

Yes. Markdown. It's about transcripts. You're absolutely right. How does Muck Rack help an organization or a company build authority through the platform?

We are kind of the gold standard when it comes to securing earned media, um, you know, measuring it, watching out, you know, getting the, getting the alerts, all of understanding the analytics around it and then the true impact and what to do next and all of that. That's, you know, very core to Muck Rack, but we also have built this AI visibility product called Generative Pulse. And, um, it's basically a citation tracking product and it's really built for comms teams specifically, not for marketing teams that are trying to like, you know, do whatever they're trying to do, but really for reputation and understanding how your brand is being cited in LLMs, particularly in the media, though you can see it across Wikipedia, Reddit, you can see every single down to the URL. What URLs and what journalists, what outlets are influencing your answers. And what you can do then is basically close the loop. Um, when you see an, a journalist, you can take an action right there and reach out to the journalist right through Muck Rack without ever having to leave the platform.

Well, Linda, this has been a great conversation. And like I said, it was ripped from my day. And I think here's our through line. In order to fight through anything, whether it's reputation or just through noise, whether it's negative noise or just noise from a consumer point of view, the energy needs to go towards authority. And there are platforms out there like Muck Rack that will help you gain and find that authority in whatever you're trying to accomplish.

Perfectly said.

Well, thank you.

Thank you for having me. The study is available. It's called What Is AI Reading? The most recent version is from May 2026, and it is on the Generative Pulse site for free. You can download it, and it goes even into the different industries too. So if you work in a specific industry and you want to help understand like who's influencing and how they're cited, it has a lot of really interesting facts about like how ChatGPT cites versus Gemini, et cetera, et cetera. So check it out.

It's cool.

Yeah, it's great.

I'm definitely going to check that out. So thank you, Linda, for spending this time with me. I knew we were going to have a good conversation and I would be able to take some very good nuggets, uh, for that. Anything that you wanna say about Muck Rack right now though? Like what, what is Muck Rack doing now that you really wanna put out there?

Muck Rack is doing amazing things with GEO, obviously with Generative Pulse. So, um, I think that's all the rage right now and people are just trying to learn about it. We actually have a course that's free. It's called Fundamentals of GEO. If you're not fully understanding, like it's a 2-hour course, you can put it on your LinkedIn if that's your vibe. Um, but what it does is really help you understand how GEO works. And how you can really have an impact. And I think like that's a skill that every communicator needs. And then we're just doing a ton with agents and media intelligence and using like our insane, like, vast data to really help our customers dig through the noise of their coverage right now and really only get insights into the stories that matter to them so they can really like go in and take a swift action. So it's all about like really highly curated kind of specialized services for big brands that need it now more than ever because reputation is so impacted now. It's so quick up and down and you can just get completely cut off at the knees. So we need those services, right? We need it to come immediately. And AI is really helping us be speedy about it. And then, you know, we have analysts to really take it from there and help you like, hey, this is what you should do based on the data. So data-driven, obviously.

It's all about the data and it is how much it affects the reputation. Linda, thanks so much for joining me, um, on the podcast.

I had a blast. Thank you.

Thank you.