Do Press Releases Work for AI Search Visibility? I Tested It.

📬 Love the podcast? You’ll love the newsletter.
 Get the weekly 3-2-1 on AI search + marketing: Subscribe

In this follow-up to last week's episode with Jonathan Bentz, Cassie Clark documents what happened when she tested one of Jonathan's top recommendations: using press releases to strengthen entity authority in AI search.

On Friday, March 20th, Cassie published a press release through EIN Presswire announcing her ebook, Freshness, Structure, Authority: The Framework for AI Search Visibility. Within two hours, she was appearing in AI-generated answers across ChatGPT, Google AI Overviews, Google AI Mode, and Gemini — listed alongside established names in the AI search space. 

This episode breaks down exactly what happened, why it worked through the lens of the FSA Framework, and addresses a recent Search Engine Journal article claiming AI search barely cites syndicated press releases. Cassie also discusses the ethical considerations of using press releases for AI visibility and provides clear guidance on what qualifies as valid press release news.

In This Episode:

• The full timeline of what happened after the press release was distributed
• Why PRLog and EIN Presswire produced different levels of impact
• How the FSA Framework (Freshness, Structure, Authority) explains the results
• A response to Search Engine Journal's article on AI citations and press releases
• Why the data may look different for smaller brands versus large enterprises
• Ethical guardrails: what counts as valid press release news and what doesn't
• How to replicate this strategy responsibly for your own brand

Resources:

Case Study: How a Single Press Release Landed My Brand in AI-Generated Answers in Under Two Hours

FSA Framework Ebook: Freshness, Structure, Authority: The Framework for AI Search Visibility

Search Engine Journal Article: AI Search Barely Cites Syndicated News Or Press Releases

EIN Presswire: einpresswire.com

Let’s connect:

LinkedIn → Cassie Clark | Fractional Content Strategist
Website → https://cassieclarkmarketing.com

Download Freshness, Structure, Authority: The Framework for AI Search Visibility:

Amazon

P.S. Is your brand losing its "Answer Authority"?

Most series A/B and enterprise brands are being "nudged" out of AI search results because of entity gaps and "stale" content. I am opening a limited number of specialized audit slots to help you reclaim your Share of Voice using the FSA Framework (Freshness, Structure, Authority).

Request your 7-Day AI Search Visibility Audit: https://cassieclarkmarketing.com/ai-search-visibility-audit/

Transcript

SPEAKER_00 0:19

Hey, welcome back to Found and AI. I'm Cassie Clark, an AI search optimization expert, a fractional content strategist, and the host of this show where we break down AI search optimization, GEO, AEO strategies, and what all this means so we don't get lost in this new wave of user search behavior. Fair warning, I have a cold and I sound a little like a goose, and I'm probably gonna pause a million times because I have to sneeze. And I'm sorry about that. If the audio sounds a little bit choppy, that's probably what's happening. This cold is hanging on for dear life. Anyway, today is a follow-up to last week's episode with Jonathan and Bince, where we talked about how brands actually get cited inside AI-generated answers. If you listened to that episode, one of the tactics Jonathan recommended was using press releases as a distribution strategy to strengthen your brand's entity authority. Now, if you've been following along with the show, you know that I do periodic updates on experiments that have worked or haven't worked. I'm always testing things. In this one with the press releases, it worked faster than I expected. Now, before we get into all of it, I do want to address an article that dropped from Search Engine Journal last week that says AI Search barely cites syndicated press releases. Because I think there's a really important nuance in that data that most people are gonna miss if you just read the headlines. Let's get into it. Okay, so here's what happened. On Friday, March 20th, I published a press release through EIN Presswire announcing my ebook, Freshness Structure Authority, the Framework for AI Search Visibility. By the way, you can get that on Amazon if you're interested, and I'll link it in the show notes for easy access. The release was written using the same principles that I teach through the FSA framework, and I wanted to see how quickly, if at all, a strategically placed press release influences AI-generated answers. Now, tiny little backstory here. I'd actually been experimenting with press releases before this. Earlier in the week, I published a release through PR Log, which is a free press release distribution platform. PR Log is fine for what it is. It gets indexed, it creates a page with your name on it, but the reach is limited. But again, that's fine. If you think about every piece of content going out as trading data, even if it doesn't do the numbers like you hope, it still is doing something. It's reinforcing your entity authority on third-party websites. Now, to be honest, when I used PR log, I didn't see a dramatic impact from that one alone, though it may have contributed to some of the momentum I was already building. Again, another piece of training data out there. Now, something interesting happened though. After the PR Log release went live, I received an email from EIN Presswire offering a free trial. It was one press release, no charge, distributed through their full syndication network. That goes to platforms that includes AP News, Google News, and hundreds of local and national news affiliate sites. I figured, if it's free, why not test it? I already had the copy written from the PR log press release that went out, so I submitted it. Now, that release was distributed roughly around 5.25 p.m. Eastern time on Friday afternoon. Friday afternoon might not be the best time to put out a press release, but I was excited. It came from my email. Let's try it, let's see what happens. Here's the timeline of what happened next. Now, by 5.55 p.m., so just 30 minutes or so after the official submission, the press release was pushed out into the wild. They got an email about it. By 6.24 p.m., Google AI Overviews was displayed in the FSA framework as a recommended approach for AI search visibility. A separate query for Recommend AI Search Strategist listed me as the first individual AI search specialist above established names in the industry. Now, we know that lists change over time. I just thought that was interesting. There are a couple of other references I'm going to mention, like, oh, I was first, but again, it changes a couple times. Now, by 632, Gemini recommended me as the top fractional content strategist specializing in AI search optimization. Again, that list changes. We know this. But it also had the full FSA framework description and my exact brand positioning reflected in the answer. Here's the interesting part of this. I have had Cassie Clark as a fractional content strategist for early stage startups nearly everywhere. But I've recently started adding in Cassie Clark as a fractional content strategist for startups and enterprise brands. And that new edition, the enterprise brand part, was included in my description at 6.32 p.m. By 7.10 p.m., Google AI Overview listed the FSA framework as a framework for AI search visibility under the heading and in quoting here, the current industry standard is defined by several core frameworks and pillars. The FSA framework was listed first. And by 7.48 p.m., a query for AI Search Expert in Google AI mode returned my name under the Emerging Voices to Follow list. Again, listed first directly after a section about the CEO of OpenAI, Sam Albin. Now, all of this was verified in incognito browsers and private mobile browsing. And the next morning, I had my sister, someone who has never searched for anything related to AI search, who tells people she doesn't know what I do, I just work from home on a completely different device, different network, different location. I had her run the same queries and the results were the same. So this was not just personalization, but this was entity authority and action. That distribution report showed the press release was picked up by over 700 syndicated sites. Now again, if we're thinking about this in terms of training data, that's 700 index pages across legitimate news domains, including the Columbus Dispatch, The Desert Sun, Knoxville News Centennial, my own local news station, WJHL here in the Tri-Cities, all connecting my name, my framework, and brand descriptor to AI search topics from one press release in just two hours. Now, I don't want anyone to walk away from this thinking that press releases are a magic bullet. They are not. This worked because of what was already in place before I hit publish. Let's break this down through the FSA framework. First up is freshness. A press release distributed through a wire service, it creates an immediate wave of new index content across hundreds of domains, and they're all timestamped to the same day. So those AI engines they prioritize recently published content. And so when you have 700 pages appear simultaneously, all referencing the same person and framework, that's a freshness signal that's kind of hard to ignore. Now, the press release didn't just announce a book, it really just created a freshness event across the web. Now, with structure, the release was written so that those AI engines could go in there and just take the information, you know, extract it cleanly. It had a clear headline and a concise summary at the top. The release mentioned the framework and it broke it down into the named components, so freshness, structure, and authority, with brief explanations of each one. It included a direct quote formatted for easy extraction. So when those AI systems look at the content for answers, they look for information that's easy to lift and summarize. We've tested this out. Structured content gets cited, unstructured content gets skipped. I kept all of that in mind when writing the press release. And then we have authority. This is where the syndication network did the heavy lifting. 700 index pages across legitimate news domains, each one connecting Cassie Clark to AI search optimization, and that brand descriptor to AI search expertise. AI engines are not just looking at a single page. They're nosy. They're looking for receipts, they're forever corroborating across multiple sources. So when hundreds of third-party sites all reference that same person and the same expertise, well, the model's confidence inside of that entity increases dramatically. But, and this is important right here, the press release did not exist in isolation. Before I sent it, my brand descriptor was already pretty consistent across my website, LinkedIn, my podcast, the newsletter, Reddit, YouTube, everywhere else that my name appears. The press release just reinforced what AI Engine already knew about me, and it didn't introduce anything new unless you count four enterprise brands that I'm still adding to places, but I do have that elsewhere, just not everywhere. That's the difference between using a press release as an accelerant versus using it as a starting point. It just amplified an existing signal. It didn't create one from nothing. So this past week, on March 16th, Search Engine Journal published an article titled AI Search Barely Cite Syndicated News or Press Releases. Now I want to talk about this because I think most people are gonna read that headline or they're gonna hear it and they're gonna draw the wrong conclusion. I will admit I did the same thing until I read it. The article covers a buzzstream report that analyzed 4 million citations across Chat GPT, Google AI mode, Google AI overviews, and Gemini. I will be glad when all of those become one product because it's a lot to say. Anyway, they ran 3600 prompts across 10 industries and they tracked where AI platforms pull their sources from. Sorry, tea break. That's cold, I'm telling you. Now, what they found is that the press releases published through syndication channels like Yahoo or MSN accounted for 0.32%, 0.32% of news citations and just 0.04% of the entire data set. Direct citations from Newswire services like PR and Newswire, they made up about 0.21% of the full data set. Now I know you just heard those numbers and I know what you're thinking. Those numbers are teeny tiny. Mm-hmm. Mm-hmm. They are. It's not even a full percent. Not even a half of a percent. In the headline, it reflects all of that. But here's the thing that's really interesting. In my experiment, the AI engines did pull directly from the syndicated press release URLs. Chat GPT tagged EIN Presswire as a source. Google AI mode listed the Columbus Dispatch as syndication partner right there in its sources panel. Bluffed into A showed up as a cited source in the Google AI overviews. Now, these were not indirect entity signals. Pulling from a third-party source like that, was the engine directly citing the syndicated pages? So, the biggest question is what's going on here? Why does my data contradict a study of 4 million citations? Now I think the answer comes down to context. The Brothers dream study focused on larger brands, companies that already have massive editorial footprints with years and years of backlinks and a deep content library. They're already well established. Now, for those brands, think like HubSpot or Salesforce or oh, I don't know, Apple, a syndicated press release is just kind of a drop in the ocean. The AI engines already have stronger sources to pull from, they already know who those brands are. But for a smaller brand like mine, someone building authority in a very specific space, my press release created a visibility event that did not exist before. When you go from having a handful of third-party mentions to suddenly having 700 index pages across legitimate news domains, the signal-to-noise ratio, it it shifts dramatically. The AI engines don't have decades of editorial coverage draw from when it comes to Cassie Clark marketing, so that syndicated content carries more weight. The study did acknowledge this though. They noted that smaller brands with less existing editorial coverage, they're probably gonna see different results. And they quoted Google's VP of Product for Search, who said that being mentioned by other sites could help with those AI recommendations. That is exactly what happened here. So my takeaway to all this, the Buzzstream data is probably accurate for larger established brands. But if you're a smaller brand, like a startup, a consultant, or anyone building authority in a niche space, well then a well-timed, well-structured press release through a legitimate wire service can create that kind of concentrated signal that those AI engines notice and act on. I posted a written case study on my website with screenshots of proof if you want to go look at that and see all the screenshots. Now I need to pause here and be direct about something. Mostly because I saw it on LinkedIn, and if I saw it on LinkedIn, I'm gonna talk about it. Because I know there are some people that are gonna hear this episode and they're gonna start thinking about how to game the system. And I want to get ahead of that because gaming the system is not something that I endorse. Could this be seen as a black hat tactic? If you use it the wrong way, absolutely. If you are publishing press releases that say something like best CRM software of 2026 revealed, with a comparison table where your product is conveniently ranked number one, that is not news. That's a self-promotionalistical disguised as a press release. And I'd argue that's exactly the kind of thing that will eventually get flagged and devalued, just like every other shortcut in the history of search engine optimization. Now, press releases they should be reserved for actual valid news. I kind of want to go over what that qualifies as because when I talked to my sisters about this on Saturday, we we sat and we had a discussion. Well, what's actually news? Now, at Podfest earlier this year, several presenters talked about using press releases as a distribution strategy. Mostly in the context of podcasting, so that is a podcast convention, but I think it applies to any brand, even those without a mic stand on their desk. The guidance was pretty consistent across those sessions. Now, they said, and then this is just standard press release information, a press release should cover timely insights, trends, and commentary that's relevant to your audience. It should use the headlines that the media would actually use, so objective, not overly promotional. It needs to be quote ready, meaning it includes the statements that a journalist could just yank from and use. And it should be story-driven. There needs to be an actual narrative, not just a product plug. Now, what counts as valid press release news for a brand? My sisters and I discussed it. Here are some examples. A book or ebook launch, like what I did, that's a real product entering the market. Um, a new research study or original data. So if you've done the work and had the findings to share, that counts as news. Um, a major partnership or collaboration announcement, a news service offering or significant expansion, a keynote or a speaking appearance at a recognized event. I did see some podcasters using Podfest as a press release announcement. That was a good idea. You could talk about an award or industry recognition, or even just a significant company milestone, like a funding round or a major client win or a notable growth metric. Now I think it's also important to cover what doesn't count as news. Now, if you have something like we think our product is the best, that's for your blog. That's not a news. That's not news. That's not a press release. We're not posting listicles where you rank yourself number one. I saw the example of this on LinkedIn last week and wanted to call it out because that doesn't count as newsworthy. We're not sharing repackaged blog posts with a press release headline, and we're not sharing anything designed to manufacture authority rather than report on something that actually happened. If you look at something and think a journalist would look at this and think this is a real story, that's valid. Report on it with a press release. If they look at it and think this is an ad, that is not a press release, don't send it. I'm gonna I'm gonna kind of harp on this for a hot minute. Now, your press release represents you. I think this is just good advice to keep in mind when talking about or thinking about your AI search optimization or your content strategy. Anything that you put out there represents your brand. Think about whether the content reflects the kind of business people actually want to work with. Because only one path, the credible one, builds the kind of entity authority that AI engines and your audience will continue to trust as they get smarter at detecting manipulation. Be the brand that earns the citation, not the one that tricks the system into giving it one. So here's where I want to land this plane. Thank you for coming along to the ride. Press releases are not a silver bullet. They are a legitimate distribution tool that, when used responsibly and strategically, can significantly accelerate your entity authority in AI search. The key is that the press release cannot be the starting point. You need to have that foundation first. You need your consistent positioning across your platforms, you need structured, in-depth content on your website that reflects your brand positioning that you're also reflecting everywhere else. You do need an existing body of work somewhere that those AI engines can reference and corroborate with. So use those clear headlines, lead with the most important information, break down complex ideas into named labeled components, and then include a quotable statement, something something that you really want reference to yourself or your brand or whoever that you're writing this for. Choose a real wire service. Free platforms like PR Log are a fine starting point. Again, training data. So anything that you put out there on a third-party website, it counts. But these paid services like EIN PressWire, PR Newswire, or BusinessWire, they offer significantly wider syndication across legitimate news domains. So the authority of distributing domains that also really matters, and it might be worth the investment. In track, what happens? After your release goes out, prompt the major AI engines with queries related to your brain and expertise. If you are nosy like I am, check incognito, check the private browsing, ask someone outside of your household to check, go to your library, your computers, but absolutely document everything to see if it's working. Again, I published the full case study on my website with the complete timelines, screenshots, and a breakdown of how to replicate the strategy. I'm gonna link that in the show notes if you're interested in trying out this for yourself. So yeah, press releases and AI search visibility. It works, but it works because of what's behind it, not because of the press release itself. Like I think if you posted something in a press release and sent it out across 700 websites that wasn't accurate and isn't referenced somewhere else, you might see that show up inside Google AI overviews for a hot minute, but I think it'll quickly disappear as a more legitimate source comes out. So do make sure that your foundation is accurate before you start putting stuff out like this. I'm gonna keep testing, I'll keep documenting, and I'm gonna keep sharing what I find, whether it works or whether it doesn't and what I think about it. That's this what this show is about. That's why I started it back in August of 2025. If this episode was helpful, hit subscribe and leave a review. I would love you forever if you did that. If you're trying to figure out where your brand stands in AI Search, head over to CassieClarkmarketing.com to get started with your AI Search Visibility Audit or book a call and we'll chat about it first. Alright, I will see you in the next episode. Until then, stay visible.