frankly… Episode 76: Building the Trust Layer with PR+AI

In this episode of frankly…

The transcript below is AI-generated and may contain minor inaccuracies. Tune in to the episode audio to hear the full conversation! 

Transcript

Dan

Hello, welcome to Frankly. We’ve got a great guest today.

Rachel

Absolutely.

Dan

So we talk with Sarah Evans, who is partner and head of PR for Zen Media. And Sarah is really kind of living at the center of AI and PR and communications.

Rachel

Yeah, she’s like AI master, but has never lost sight of her roots in PR and that job, like she’s combined them incredibly well.

Dan

Yeah, so we talk a lot about kind of the intersection between these two things, how PR can help companies with their AI strategy, what measurement looks like in the world of AI. The tools that she uses and loves. So there’s a lot here. All of these things kind of how they cross over and where they fit together.

Rachel

And I want to note here, I tell her during the interview, but if you don’t follow her on LinkedIn, give her a follow. She has incredible thought leadership that is really insightful. She’s just a really good follow if you’re in anything related to PR.

Dan

Yeah.

Rachel

So enjoy.

Dan

Hi Sarah, welcome to Frankly. Thanks for coming on.

Sarah

Thank you so much for having me.

Dan

Of course, yeah. So we always like to start with some intros here. So talk a little bit about your career path and what you’ve done over the years that have led you to Zen Media in your current role.

Sarah

It’s such a big question knowing that this is my 21st year in PR, so I’ll try to keep it as short and sweet as possible. I always tell my younger peers and cohorts that I essentially worked myself out of a safe study job. I did the college thing, went corporate America and had a great job, but realized when emerging Twitter and social media back in the day was emerging, I was moving per minor much too fast, and was then moonlighting at night consulting companies that were ready to explore emerging technology and their comms plans and PR plans. And at some point that started to outpace my day job. I was in corporate comms, government relations, head of PR, all of those things, and then started my company at a whopping 28 years old. And then in eight months ago, when it was about 16 years old, it was acquired by Zen Media, where I’m now partner and head of PR, and it’s been an amazing journey.

Dan

That’s awesome. Congratulations.

Sarah

Thank you.

Rachel

I love hearing about agency, like starting an agency and like especially to do it at 28 is really… Really impressive. So congratulations on that.

Sarah

Thank you. Pre-kids, pre-knowing what I know now, because I always joke, I’m like, I would never do that now.

Rachel

Sometimes knowing too much is of a detriment, right?

Dan

Yeah, holds you back.

Sarah

So true.

Rachel

Good for you.

Dan

Yeah. So today you’re a partner and head of PR at Zen Media, you said. So what does that entail for you? What does your work look like? Maybe a loaded question?

Sarah

Just like in all things true to PR, no two days are the same, but when you’re a partner in an agency, you’re obviously beholden to revenue goals and all of the big, audacious goals that an agency has to run, but also as head of PR, because I love what I do, I get to rebuild a department of what I believe the future of PR is and strategic comms. So it implies everything that goes along with that, and re-examining every tactical workflow imaginable, because I’m trying to find ways where AI can impact that or maybe touch something to either optimize it, streamline it, or help our practitioners work better. So if nothing else, I’ve gotten more in the weeds so that I’m not out of touch with how the industry is performing with the context of B2B enterprise and how we can do it bigger, better, and better.

Rachel

I’d love to hear your overarching, this is going to be another big question, your overarching kind of philosophy on the intersection of AI and PR. I’ve actually used your LinkedIn as an example for some clients on how they should be doing like thought leadership.

Sarah

Oooo.

Rachel

I think you do it really well. So they’re in a completely different industry, but it was a good example of like, if you want to be a thought leader online, like this is a really good example of how to do it and the kind of like content you need to be putting out but a lot of your content lies there. And I think it’s like the big question mark, as so many people are saying, AI is going to get rid of so many jobs. We know that PR is a lot of still relationship-based practices. You have to know people and have that face-to-face interaction. But how do you feel that the intersection of PR and AI, like where are those, where are you seeing a lot of it? Where is it beneficial?

Sarah

So here’s the least scary answer, because I am not doom and gloom with AI, especially with PR. And I kind of joke with my peers right now that PR is having its moment. You could know nothing about AI and keep working with integrity in PR and actually be touching what AI needs because the most trustworthy content is content that is authoritative, timely, insert your words, and most importantly, human. Because once all of the AI slop gets removed, it’s that stuff that’s going to matter. So when you’re watching my thought leadership on LinkedIn, and by the way, I believe in abundance, so even if it was someone in a competitive field, like more power to them, like let’s all create. That is my working living case study of what I believe will start to move the needle on things. And once I’m the case study, we can also replicate that for other clients. It becomes standard and best practice.

Dan

And I saw something that you had posted that had a quote in there that was media is becoming the new metadata. So not just on the, not just on the side of like AI for use for PR practitioners, but on the flip side of that, like how does media relations then kind of feed the AI strategy for some of your clients or in your own work?

Sarah

Dan, asking the hard-hitting questions. I’m looking at that technical layer, and I don’t know if it’s too early. I was going to talk to you all about what I call the trust communications layer. And I think that is now foundational and fundamental in everything we do. And it kind of hits at what you’re saying. Whether it’s earned media that you’re pitching, whether it’s owned media that you are putting on your own side, or maybe it’s social, it all has to have this layer to it. So any content. Let’s say you get a comment in a great trade publication, or just a comment in a larger article. You can now take that and make it into an own piece, much like we would suggest before. But now you’ve got to follow what I’m calling this trust layer. So when you’re creating the content, it’s topical authority. You’re writing your headlines to be shaped like prompts. Relevance, that’s that. schema markup, consistent entities to make sure it checks those boxes. Then useful proofs, I’m calling that metrics or evergreen links. It could also be things that your brand or company has done before that you need to keep repeating. So it’s training the system, the machines. Then it has that search equity layer. So you want to embed phrases that you need to repeat, test that in LLMs, and some could say still maybe an SEO layer so that you aren’t missing out on the search. And then timely reinforcement where you can loop in recent events awards, certifications, things that oftentimes were not newsworthy, they got left to the wayside once they came and went, they’re there, you have another chance to reinforce that. So if you could put all of those things and train the machines, at least what I know to be true today, that is the long-winded answer to your question.

Dan

No, I love that. And it gets back to like the, just the core of integrated communications overall also in that you, you never use one thing in one place. Everything that you do has to be in multiple places, bring in other aspects of your strategy and your campaigns and things like that and just kind of, instead of just training an individual audience member, now you’re training the overall AI models as well.

Sarah

Yes. And there’s nothing rocket science about this. In theory, PR should have always been operating like this, but we totally missed the mark on the SEO train. I knew and still know, I’d say just enough to be dangerous. I knew and know it’s important and how we’re linking to things and we want to know what SEO keywords people want to rank for, track. But now we are on the ground floor of the next lever.

Rachel

Yeah, and I think what’s important about what you said too is like, it starts with the human. And we always say that. You still have to write content for a human because it’s the human that’s doing that searching. They’re the ones that are putting those queries out there. They are the ones, right, you can back that up with the tech, with the schema markup, making sure that it’s crawlable, it can read it, it’s training the model. But at the end of the day, the content you put out there still has to be written for humans. That is like top of the line.

Sarah

And that’s why I called this a layer, because I don’t want it to be a workflow. I don’t want it to be something where someone’s like, I just got GPT AI slop, and now that’s going on my website. And I actually did make a GPT that I’ll share the link with to you guys so people can run their content through the trust layer. So you don’t even have to know, I mean, you can train yourself, but I’ve done some of that heavy lifting.

Dan

Oh, nice. That’s awesome. And yeah, I think. Rachel, going off what you just said, like the human piece of it too, I almost feel like that’s more important now.

Sarah

It is.

Dan

That AI is more prevalent, which might seem kind of backwards, but if you have more people using speech for search, you have more people just typing in informal queries and things like that, there’s not as much of those like… what you would think of a classic search term. Nobody’s, using boolean on ChatGPT. So it’s like, it’s you have to think about power person search for it.

Sarah

Remember when you discovered boolean search, you’re like, oh my God, I’m like a power searcher, now there’s power prompters. And now you’re like, now I know that I have to actually think like how people are doing intent-based search. Like do, perform a task for me instead of what is something?

Dan

Yeah. So you just talked about the GPT that you created. What are some of the other things that you’re finding to be like some of the more valuable use cases in your daily life for things like AI GPTs?

Sarah

So many to mention. I’ll just tell you a quick personal story. I had my oldest just started high school this year. And I had four moms of other freshman year boys who gathered at my home Friday night. I said, bring your computers. I’m like, we’re going to make a GPT to help our kids become more efficient with their homework. And we sat here and built our own GPT so we could feed in their schedules, their learning work styles, you know, everything and gave us a better output that then we connected to We’re using a task management platform for the kids that connects to their Google cals, so they’re breaking down all of their tasks in time increments that work. Take that times 100 for everyday workflow. And I’ve done that even with my own career and our teams, like here’s how to be more productive with your calendars after you are on a call. You can take that AI software, feed it into your task management platform, AI integration to look at how you work most productively, break it out into tasks. What used to take me two hours, in essence, now takes me two minutes, and that’s just one piece of the pie. We are not using it to completely automate all of our content. Like people who see my content online, they’re probably like, well, she’s just… churning out AI-written press releases and it actually is not. We’re doing what you said, Rachel, we’re actually going human and then layering on things to make sure it’s going to perform well. So we’re still doing work, but we’re able to do more work better.

Dan

Yeah.

Rachel

Yeah, you are filling in where it’s important to be human, right? There are so many things that AI can do, not even better, but quicker, and just give you the time back to be able to strategize and use your brain where it needs to be, right? I mean, that is, if people can’t understand that about AI, they’ve missed the whole point. I feel like you’ve missed the boat.

Sarah

Yes, and I tell my team, and I do this intentionally, I keep my AI detector pulled up, and I know there’s controversy on how accurate they are, but if anything comes back, more than 2% AI generated, I tell the team that’s too much and you’re going to get a full rewrite for me request.

Rachel

I think it’s content and then even just like as a consumer or the practitioner using that. It’s really a trust layer, I mean, that’s a great way of putting it, but like having those goggles on that you can always see. I mean, how many times do you get duped by like a video on TikTok and then you click open the comments and it’s like, this is AI, like AI spotted. I’m like, whoa, like you really can’t believe anything anymore so when it comes to the work that you’re putting out, especially when it’s on behalf of clients, I think it’s even more important because they have that trust with you that you aren’t going to, you know, put something out there on their behalf that is far too AI so I think your 2% is great. But maybe talk a little bit about how you talk to your clients about your use of AI and what you’re doing as a company and how it’s making you better practitioners.

Sarah

I think all of our clients are well aware that we are using AI, but we tell them how intentional we are. We’re not sharing anything that’s confidential, proprietary or similar in any of our tools. We’ve actually built in-house one thing that I think is really unique about Zen is we have 5 AI developers on staff, and we’ve built some internal tools to us that are SOC compliant, all of the things that we need to work, which is really unique and interesting. When we’re talking to clients, though, ironically, even if our client sends us content, I run their check, I run their stuff through Gen. AI Checker to make sure they’re not giving me content to make them look bad. As far as disclosing what we use, I think it’s much like we don’t say, unless they ask, we’re not saying, oh, we use Business Wire or PR Newswire, unless they specifically ask, we’re using Distribute, but we’re using all of the likely players, but also they’re all integrated into our internal flow so we can turn on which platforms we’re using. So I’m not actually out in browser land toggling between services. And then we use ClickUp as our internal task management platform, but that has an AI layer too. So that gets smarter as you use it based on what you’re creating with clients. So it’s kind of at discretion, but if I had to offer some sort of like general blanket statement, it’s yes, we use AI but we’re not putting out AI-generated content.

Rachel

I think that makes sense. I think it’s also the types of clients you’re working with, they’re more than likely using AI for things too, right?

Sarah

Yes.

Rachel

So it’s, and as they should, I mean, they could probably get different use cases for it, but as they should.

Sarah

Yeah, we’re constantly having, Rachel said you’re on your AI task force. We have one internal and we’ll have actual meetings even to see how could this impact the sales process? How does this impact what we can do for teams? Let’s nail down this process to a T to see where it could be automated in some way or where it still needs the human layer. We’re looking at things like that and more to I mean, I tell my team, if you want more time with your family, you want more time to breathe. Sometimes they can take care of those things without impacting the quality of what we do.

Rachel

Yeah. I think one big question that is floating around, at least like something we’ve talked about and we’re still kind of diving into, but I want to talk about measurement, right? So understanding the impact of how you’re using AI with PR and how are you measuring that to say like, this made an impact, right? So you’ve You’ve got the media relations, you’ve got the coverage out there, you’re using it as an owned piece. What are those KPIs? How are you measuring it?

Sarah

So we are in, I call it an infancy stage, but I think much more ahead than most places. When I say we have AI developers on staff, this isn’t public yet, but it’s okay. I think if I share with you guys, we do have a proprietary Gen AI visibility tracker. So we’re pulling baseline for all of our clients right now. And we monitor mostly on a quarterly basis, just because as we’re all learning, how fast are these being changed or impact? And we do see fluctuations on a weekly, sometimes daily basis, but where is it? Where is it leveling out on a quarterly basis? We’re looking to see right now, we can actually pull what prompts ICPs are looking for, and those are what we’re going after. And that’s where we’re reverse engineering our quarterly strategies against and we might have a dual pathway approach. If a brand is like, I’m not fully, our leadership isn’t fully on board with Gen AI, so we still need, you know, traditional media relations, we still need authority, reputation, you know, building through one pathway. We’re also seeing if they’ll add on dual pathway where we say, can we, let’s see if we can impact these prompts, ones you weren’t showing up before, now you are. Your competitors were showing up here before, now you are. And we’re looking across the board here. But also sometimes just the value in pulling those reports is really important to them and it’s actually impacting some of their sales strategies. They’re like, I had no idea this brand was being flagged as a competitor in AI. I had no idea our ICP were actually searching in this capacity for what we do.

Dan

Yeah, it’s almost like share a voice in a traditional media report,

Sarah

Yes.

Dan

But just for a new kind of source, a new zone.

Sarah

Right on, Dan. I mean, it’s again, not rocket science. It’s like, here’s another, just recalibrate what we’re doing in this format now.

Dan

So talk a little bit about kind of what excites you for the future of AI. What do you see as maybe like, or what’s a use case that you would love to see come out in the future for your, either for your own work or for client work?

Sarah

I mean, I am very much an optimist and I like to consider myself a first mover. I get really excited about things. I just feel right now the same energy I had circa like 2011 Twitter when things were new and exciting. And I had launched JurnChat, which was the first industry chat on Twitter. And I ended up traveling to CNN and doing the first live stream there with the CNN iNews team and going to NBC and doing the same thing. And I feel like we’re in that place right now where what we’re creating and sharing is really changing the industry and how we work. And the reason I’m excited is because I believe there’s more attribution for what PR does. For so long, we’ve been the soft metric. Here’s our, you know, estimated reach, unique views. And it’s like, what does that really do? We can’t prove that article resulted in a sale. And now when budgets get cut, we’re one of the first ones slashed. And now I feel like That’s probably what excites me most. We can attribute the more we can track what’s happening in generative AI visibility to the impact we are creating.

Rachel

I think it’s also a good point to talk about communications versus marketing and how it gets lumped together alot. But at the end of the day, communications really is still that like general brand awareness, getting out there versus marketing, really focusing and working with sales, right? And trying to get leads. Like there is a differentiation and I think there is more discussion around that. If for so long they were fully separate, then people lumped them back together and now we’re kind of saying like, they do serve separate, like they have to work together, but they do serve separate functions. So hopefully with what we’re doing right now in PR, we can kind of say like, we also still have, like we might have some impact on the sales, right, in PR.

Sarah

Yes.

Rachel

But realistically, We don’t have to. That’s not really what communications and PR is existing to do, but we can, and now we can measure that.

Sarah

And when people can start purchasing, and they can in some cases make purchasing decisions within the LLMs that they’re using, and let’s say we now impact an answer in a prompt and someone makes a purchase decision off that and we can track it. That’s huge. And I always used to say that marketing is what got people in the door and PR is what kept people coming back. But now I’m seeing this inch into that getting people in the door through the answer pathway.

Dan

Yeah, that initial research phase is so important because you know, the first thing you type in, if you can’t be the first one that pops up in that answer, it’s you’re already out of the game almost at that point.

Sarah

Yes. And now’s the time to get content out. I just also, not on the doom and gloom side, but at some point I do believe that this information funnel will either be closed off or limited or have guardrails on it and right now we have a big opportunity.

Dan

Yeah, so we were talking about that a little while ago, actually, Rachel and I, of like, when does advertising start to come into play on GPT and these types of things? And at what point does that just completely, again, change the game and make it, make us all rethink and re-strategize and look at where we’re spending our dollars and things like that.

Rachel

Yeah, that’s an interesting play for like your trust layer, I think, Sarah. Like that to me removes a layer of trust with these models when we start putting advertising into them.

Sarah

But will they try to be like the viral, like deep fake on TikTok where like I’m going to use the trust layer in my ad? You know when you’re searching on Reddit and you get interrupted by an ad, and you’re like, you totally do not get the community here.

Dan

Yeah, this is not the right spot, right?

Sarah

Like, what were you thinking? So, you need that trust component to ads, and maybe those ads are truly inserted in the right place somewhat organically. Like, oh, you were searching for this anyway so here’s the person who paid the most to reach you.

Dan

Right.

Rachel

It’s an interesting way of looking at it. I’d actually love to hear your take on Reddit’s impact on LLMs and generative AI, because it’s something that we, I think we’re seeing so much more and it’s not organic for companies to show up in there. So how do you utilize your stakeholders or third parties that really believe in you in these communities that then impact how you show up in generative AI.

Sarah

Funny enough, we’ve got several of our brands looking to now execute on a Reddit strategy. And we’re doing it. I mean, I’ve been on Reddit for 17 years. I don’t even know how long. But we’re very mindful of what moderator guidelines are and how subreddits work. And most of our clients that want to be on there are also Reddit users themselves. So they’re well aware and they’re not doing it from a paid capacity. They’re looking at how do I integrate in the subreddits where we’re most likely to provide value and either answer questions. A lot of them are taking the vantage point of how can I be there to be a source of truth to answer questions without being blatant promoters. And we know that even writing content in a question-based format helps now train LLMs if they’re going to be using those Reddit sources of, as sources of truth.

Dan

Yeah, and outside of Reddit too, like using those, using the questions that come up in Reddit as guidelines for your own content on blogs or whatever it might be, like that’s such a powerful way to kind of use your user’s own words to kind of shape what you’re putting out there to be helpful even on the platform.

Rachel

It’s truly a peek behind the curtain of your customers and what they’re looking for and the questions they’re asking so if you are looking for more content to put out, it’s a great place to look in those subredits. and understand how to use. Even if it’s, we’ve talked about this from a customer advocacy standpoint, do you have some customer advocates that really believe in your product that are in there talking about it and it feels a lot less promotional to do it that way too? You know, like how do you use people that already love what you’re doing and have them get the word out?

Sarah

We have a client that has like rabid fans and they have their own subreddits. And it is amazing. It’s like the gold best in class standard of what happens when your customers become advocates. And when we want to launch a campaign, it’s never like, will this work? It’s like, this will sell out in two hours so how do we want to leverage it?

Dan

It’s a good spot to be.

Rachel

Yeah.

Sarah

They nailed it. They’re not even running it. It’s their community.

Dan

Yeah.

Rachel

I feel like I’ve seen that done really well, even with like Oura Ring is an example of one where I’ve seen it. The Oura thread, they have like blank at Oura and it’s basically like their customer service in their answering questions within their subreddit. It’s not, you know, it’s not promotional, but I think they’re a really good example of doing that well within their own subreddit. And I’m sure that makes a difference for how they’re showing up in LLMs for sure.

Sarah

Because it’s trust. It’s trust.

Dan

It all comes back to that.

Rachel

It does all seem to come back to that, Sarah.

Sarah

It does.

Rachel

I love that. This is kind of… I know you guys have your own tool, so maybe this is a loaded question, but if maybe what is your favorite widely available generative AI tool to use or where should people start when they are like coming into this and looking for something to help them improve their processes?

Sarah

I’m always okay talking about tools. That is one of the things I am an absolute junkie of, but I’ve been for years. I will beta test just about everything. I’m on product hunt every day. I love testing out new things. If you’re looking for a way to do generative AI visibility paid, the best one I’ve seen today is profound, and they I think just got their series B funding. Even then, they’re limited on the number of prompts that you can run. So when you, anyway, I won’t go down that path road, pathway too far. GenSpark is huge. I love, and now Claude has launched Comet, they’re a competitor to that, it’s like your super AI agent. So if you’re looking for something that kind of does an all in one, I’d highly recommend that one. And then a third one, just to help expedite on visuals or anything that you’re creating as a visual component, Gamma app. Genspark’s kind of challenging them there now also, but I’d say that those are my top three.

Rachel

I love that. Do they have free options or are they all paid?

Sarah

I pay pro for every single tool I have. So I think they all do have a freemium, like you can run one or two things to check it out, but you definitely have to pay. GenSpark especially because it’s so valuable.

Rachel

That’s fair. If you’re, hey, if there’s value in what you’re paying for, people are just generally willing to pay.

Sarah

My time is valuable. So anytime I get it back, it’s worth it to me

Rachel

That’s very fair. I really am interested, Dan, if you have something else, please.

Dan

No, go ahead.

Rachel

But, I am super interested in one of the questions that we like to close out with, because I feel like you’re a really good person to ask this to. Knowing what you know now, what advice would you give your 21-year-old self?

Sarah

My god. It would be nothing to do with tech and it would be all about… knowing your own personal value. If I go back and tell myself, I would say, remember when you are interviewing for jobs, that you’re also interviewing them. Because finding a place where you have purpose and you feel excited and you feel like you’re contributing is with a caveat, more valuable than any paycheck, but obviously you need to make money. But I have seen what happens now being in the industry for so long. When you wake up and don’t want to go to work, that’s a very heavy mental toll. And I would say that emotional fulfillment, that purpose-driven fulfillment is much more valuable.

Dan

Totally agree. Both are important, but that if you don’t have one then, or if you don’t have that one, then it’s, it can just be miserable.

Rachel

It’ll affect every other part of your life.

Sarah

It’ll force you to start a company.

Dan

Yeah.

Rachel

So that’s why you started the company at 28. Okay, got it.

Dan

Going full circle here.

Rachel

I love that. Dan, do you have any other closing questions?

Dan

No, just thanks so much for coming on. This was great. Love talking AI with you. And yeah, hope we can do this again sometime.

Sarah

Yeah, thanks guys. Great questions.

Rachel

Yeah. And if you’re not following Sarah on LinkedIn, we’ll put, we’ll link to your LinkedIn in the description because you’re a great follow, so highly recommend. I want to make sure people, I’ll plug that for you. don’t even have to do it yourself. Thank you. Thanks, Sarah.

Dan

Thank you. Well, thank you again to Sarah for coming on. Really enjoyed talking with her about not just PR, but all things AI here also. So definitely check her out on LinkedIn. Give her a follow. Keep up with all of her all the things that she shares, there’s some really, really interesting things about how these two worlds are intersecting and overlapping in a lot of different ways.

Rachel

Yeah, you might be looking for a solution to something, and she probably has the answer, whether it’s in a tool or really just how she’s using something. So highly recommend, and we will see you next time.

Dan

See ya.