Part 1: AI, Marketing, and Cutting Through the Noise with Yael Macias

Welcome to Promptcast, your go-to podcast for AI, security and everything in between. Join us as your host, E-Tamar GoLon, CEO and co-founder of Prompt Security. Dives into the latest trends and insights. Let's get started. Hello everyone and welcome back to another episode of Promptcast. I'm your host, E-Tamar GoLon and we have a very special episode planned for you today. I'm joined today by someone who's been a very special part of Prompt Security since our early days and has played a major role in shaping the journey, actually the brain behind Prompt's marketing and even this specific current podcast. So it's actually our coming full circle. Please welcome not only the sales of the sales of Prompt Security, but also the sales of the company. Thank you for joining us. Hi, E-Tamar. I can't begin to explain how strange it feels to be here after we launched this podcast a year and a half ago. I did not think I would ever be a guest here. But yeah, it feels great to talk about all things AI and marketing because I wasn't going to be talking about the security part of things. It's definitely an exciting podcast today. How is it to come from the backstage of orchestrating everything at the backhand and now talking in front of everybody? I think it's kind of symbolizes part of our special journey here, isn't it? It does. And I mean, you obviously know me by now. One of those marketers that I like not necessarily being in the front or doing, not talking so much about what I'm doing. So it's been an interesting exercise to put together some notes to prepare for this and definitely very much a full circle moment. And I mean, yeah, not to be too cheesy, but also what a journey it's been, right? The past couple of years. So we're also recording this at the end of 2025. It's been definitely an interesting year and lots to look forward to. So yeah, let's dive in. Cool. So let's dive into that. Actually, I was keen on having this episode. I'll tell you guys when Elan and her team spoke to me a few months ago about candidates to be a guest on the podcast. I thought that I really wanted to have an episode where we discussed the impact of AI over marketing. Maybe say it's go to market by an old implication in the last couple of years since AI became so substantial in every domain of our life. Both, of course, the good, but also the bad in the ugly. So if we usually talk about intersection of AI and security, today I kind of want to talk about a little bit different area in the peripheral domains of AI, but this time the intersection with marketing, AI, for example, as a tool, AI in messaging, in content generation, AI as a new go to market motion and you get you get a feeling. Last season, we had a series of questions for the fire around at the end. Hopefully you remember. The one of them was where were you when Chagy PT was first launched to the public in late 2022? I say, "I'll start with this." And throughout the episode, maybe start building in a chronological order, how it was impacted marketing teams from that moment, like two, three years ago to today, right? Sounds good. As millennials, I feel like we have all of these historical moments that we've lived through unfortunately, most of them tend to be like bad, right? Where were you when this happened? How is the pandemic for you? But I feel like the public release of Chagy PT kind of makes it to that like landmark or historical moment that as well as millennials, because we like to make things about ourselves, but as this generation, like it's definitely a before and after. I don't think at the time we knew what we had in our hands and to this day, I think it's still evolving and it's still changing. But actually, it was curious because I was at the time at my previous company. I was part of a phenomenal marketing team. And I remember seeing a Slack message from one of my dear colleagues that she wrote on the marketing private channel. "Hey, I have a cousin that's working in this company. It's called OpenAI." And they're like releasing this thing to the public in the next few days. Now it's kind of like friends and family, some demos, seems like it's going to be pretty cool. I don't know. We should like think about playing with it. That was a message. Like, I think we all get lots of messages like that all the time about different things. So why is this different or why is this special? Fast forward. That was maybe a week or 10 days before the public release of Chagy PT tool. Yeah. Well, to everyone. And even in the beginning we started playing with it mostly as a joke. We're not technical people. I used to play with it with my friends. This is something you think to write love poems or mockery to friends, to family and kind of like an entertainment tool. But very quickly, it obviously changed. At the time, the team was growing and someone had joined the team fairly recently. And one of the initial tasks that I had given this person was to take a bunch of different content sources and reassembly to put it together as a blog for contents indication. For me, this was a task, especially a new person that would have taken, I don't know, at least a week, including reviews, consultation, back and forth. We didn't join it for hours. I had on my inbox a draft of this thing I had asked. And I was like, wow, that was fast. Not so cynical anymore. Aren't you? Yeah. Wow. That's interesting. But for me, the interesting thing is that at the time, obviously this is very early on, people aren't necessarily talking about the fact that they're using AI. It almost felt like cheating. And these persons told me, well, yeah, I used Chagy PT to help them. Oh. Okay. The first thing that exists person like confessing to doing something that like, you know, this is not okay. Like you're using a tool that's not acceptable to use to create content for work. We're talking about January of 2023, right? This is like four weeks after the public release. And then very quickly, I was like, actually, this is just, okay, this is brilliant. I mean, this person has just kept through the entire onboarding. It's straight on delivery for content creation. And obviously from there on, it was just, well, skyrocketed. We could produce if we wanted blog posts on a daily basis. Another very early use case was taking it a step backwards, right? Like not even giving existing blog posts or content sources. We had like this webinar again. I'm talking super early days, right? So we can then write like, feel where we were and where we're at today. I'm not sure. Yeah. I'm not sure. But we'll, we'll roll with it. Maxon will have a couple of episodes. And there was this person that was supporting with a campaign. Wasn't part of the product marketing team. Not someone that was supposed to be an expert in like content writing, nor about the subject that she was asked to write a blog post, which was just supposed to help like, oh, where do I even get started? We have this webinar. We need to put together a blog post. I'm like, why did you get a transcript from author AI? At the time, it was called author AI. I'm pretty sure there was not a lot of AI into it. More so like a transcription. In memory. Yeah. And like within like 24 hours, this person took the transcript of a webinar with a few prompts said, okay, give me the five key takeaways. And there it was. A blog post, someone that would have probably never thought of like started to write blog posts at the time because it wasn't part of the job description. But hey, it opened a brand new world. So yeah, fast forward. You're sitting in meetings with a go to market leadership team. How can we use AI more in our job? Like, no, like that's not how it should be. Should be. First, let's decide what we want to achieve and then gradually incorporate AI into it. But of course, I'm sure you also sat in a lot of these meetings. I should be using more AI. How do we do it? And salespeople using it for prospecting. Suddenly they're prospecting because it's cool because somebody's writing emails on their behalf. Anyway, we can talk more about that. I think at the time, and this was obviously early days in parallel. There was also the efforts from R&D on, okay, how do we actually incorporate AI into the product? Like so many companies and items. Of course, you know, this is well because of you. You're fast life. But how do you actually use AI to make something in cybersecurity better for remediation, responses, just recommendations to give some additional context into something? That stands for like at the time where yeah marketing and sales are play with these to write better emails on blog posts and in parallel are in these like building features on top of on top of LLMs like that that that are delivering better security outcomes like something that was growing and and yeah here we are today where my mom who's the least tech savvy person and I love her and she's brilliant but she's someone who embraces technology fast. I was standing everybody and actually it's true it's not a lie but doing like fundraising and pitching to customers I told them a real story that my 90s something old grandmother years old grandmother she was using Chad Gbt for making a chicken soup so yeah she's stealing my joke and building some analogy with with her mom. Yeah well I think yours still like I think yours is still married like the attention it got at the time and I've definitely heard that joke before but well no joke it's actually a real story but yeah in my case for me it was like a realization when I first saw my mom use a Chad Gbt to look for something like usually for a coffee shop like it wasn't even like something to anything but yeah it wasn't Google I was like oh it just just looked for it up on Chad Gbt like wait what like this is way beyond the technology with like the high tech bubble where I mean this is not marketing this is not R&D this is just people incorporating this into their daily lives and I think for me that was like the realization of okay we're onto something here so I was a very long answer to a question but that's a whole point of having a question please answer with yes or no okay no no actually I love the I love the answer the question was was a tough one I don't think there is a short one to to deliver here it is indeed a long journey and I can share with the audience that yeah although all the all the setup of leading the marketing in a talk tier up fully security for a eye company she was a little bit cynical she is still she's not like a complete fanatic fan like myself she sees the inner perspective the issues in that and where not to use it I think it's a more balanced perspective I think it's going to change everything in the next like few months so I'm on the other radical side but we are definitely going to come back to how you use AI in the day to day but let's flip the script for a second if I may today every company is AI powered AI driven AI leveraged I see AI everywhere even when I'm washing my car or going to the dentist it's written like powered by AI or something in that spirit which is becoming a little bit of cliché and between you and I it's we cannot see it anymore because AI is real there is a real substance behind that we do not need to push it into everywhere with no reason so when everyone's like when everyone sticks AI into their messaging their content what's your take on that is that bad good in between depends how do you see it yeah you know one of the two things that characterize me one is that I'm very cynical and the other one is that when I don't want to give a straight answer I say you know what I'm torn about this that way I don't have to like bear the responsibility of having a bold statement but the reality is that first of all it has changed a lot over the past five years obviously right everything that revolves around the AI but the but AI had its moment in marketing before charge of it was a thing like there were a lot of companies that were talking about AI as like this premium feature something that was accessible only by a few the AI teams within companies were like the most elite people and there was something like very niche and very sophisticated about having AI as part of the product but with that also came the lack of explainability like a lot of companies especially in cyber security which is obviously what I what I'm most familiar with started talking about AI and how they're incorporating it but it was so hard to understand from the customer perspective that at some point it just became like wait so your product is a black box like you have some AI doing something but I'm not supposed to understand how it works so it went from being something nice sophisticated premium feature to this is to complicate it people are very skeptical about this let's just not talk about AI anymore I actually haven't anecdotes from from my previous company where we had a domain that was dot AI and we did a migrate migration of the website to dot security now with insight bias it's like I maybe we shouldn't have changed it because then AI had its moment I'd still work the was a point on time that a company or a CEO or a board of an organization wanted to stay away from AI yes but I'm so ridiculous today it sounds ridiculous to think about it now like we're past that curve of the excitement I mean this is not a hype cycle of garden I think we have to talk here about the hype cycle is you guys are familiar with the shape this is more of a roller coaster of like changing features and an excitement on on a daily basis almost but I still stand by it like now with everything that's like past I don't think it's been a few years since this move happened from like AI to security forget it's just an anecdote but the product marketing person in me tells me it doesn't matter how you're doing it at the end of the day the customer cares about what's the outcome is there's an AI embedded into it great explain at some point like how the solution works like he doesn't care if you wrote it in Python or Java if it's in the cloud or on prem if it's freaking work give it to me yeah and and have some explainability which is something that now we have five years ago we didn't have and that's why it also became something so inaccessible and so niche but there's also the skepticism that comes with talking about AI where so many companies and this is like brother problem and just a messaging thing so many companies have emerged from just being basically a wrapper on top of tools that already exists like on top of chat with you or or or or or or or other tools so yeah like the skepticism there is because okay I can achieve the same thing by using chat with you like what's the story here why do I why should I care about the AI that's there so it there was it had its moment then it stopped having its moment then it had its time again when chat GPT became a thing like the opposite of the hype cycle what you describe like the same graph but symmetrical yeah and now it's like just don't talk about it like it's there it's obvious if you're a B2B company and like that wants to survive in 2026 of course you're incorporating AI to some part of the product like that's almost a given just sure what are the outcomes and try not to scare people off by saying that everything's going to be processed by AI and you're not going to know and you're not going to have the explainability like I think humans still want to be yeah actually it's becoming the normally you always want to be special and different so today maybe we are going to see gradually more and more messaging of powered by human learning I based yeah not AI powered fast platform for final optimization what does that even mean I remember when I first started talking when obviously to you and also to to Neor or founder and and CTO prompt one of the things he told me and as a technical person that at the end of the day it's also the audience were trying to to to reach saying I just don't understand like what anyone is trying to say anymore like what are all these websites everybody's talking in a way that's also theory like what what are you even saying what do you sell what's the product so I think AI has only like adding infusing AI to the messaging has only made this worse and and and last clarity I think many companies saw for a second that what if I just say that there's AI included will sell better and then there was this phenomenon called AI washing which is just add AI to everything and say it's powered by AI even if it's not necessarily true like there's just some degree of automation to it and you're calling it artificial intelligence but I do think we're also passing that especially again in in B2B and there's another nuance there and it's core you selling to like an example and and I'll bring it back to cyber security again because obviously this is what what we're the most familiar with and also to be a bit more in line with with the content of the podcast but now there's so many companies talking about automation at the so okay I agents to make try to do an investigation easier if you're selling to a seesaw then yeah definitely say that there's AI agents that are doing the work and in three years from now you might be able to replace like half of your tier one analyst because savings better outcomes board is happy why not now if you're trying to sell bottom up or you're trying to engage also the tier one analyst and you're saying that an AI agent you're supposed to train regarding his face exactly like like, yeah, in two years from now, this agent is gonna, we'll have learned everything of how the workflow operates and then eventually replace you. So I think a company called Seven AI is doing a nice job there because that's what they do with AI agents to automate security operations. But they're also talking very much about doing human work. And that way everybody's happy, yeah, you were since security analysts, who human work don't do things that can be uploaded, but yeah, your job is yours. You just should be focusing on other things, whether that's through in the long term, how does that sustain? That's a different story. But I think a word of caution there for everybody that's using AI agents is like that, that also make sale-earned, the ultimate selling point might be good for the sea level, the person that at the end of the day might have to compromise on what they're doing for a living, that's a different story. Yeah, I tend to agree. And yeah, let's be honest, I guess that some of our listeners right now came to ear from you. The magician of prompt about all marketing strategy and tools and how did we do it? Apparently, I don't know why prompt is considered like as a big success of marketing. I'm always telling them that this is purely myself and my brilliant mind. But if we admit that you have some portion of it and we want to argue about the exact percentage, I want to ask you three questions about how do you do it? Which tools do you use? Like actual concrete recommendations for market years today? So of course we should start from AI, I will not. And therefore the question, how do you use AI today for marketing? For almost I think 18 months you operated as a solo marketeering prompt. Though hopefully it looked like we have an army of marketeers. It was only a L and myself on the side giving some non-creative ideas. So you don't have to reveal all the secret sources, of course, but would love to pick your brain about it. Yeah, and yeah, we can agree that you have at least 95% and I'll be happy with the 5%. But-- Neither very good negotiation. No, and also like, you know, I'm fostering from a very multi-cleaned negotiation. Let's agree on that. Fair enough. So, okay, so few things. Not everything was AI and not everything is AI. We can have a separate podcast with content and messaging and all the things we did that were very much a team sitting in a room and having ideas and thinking creatively. And there was also a lot of manual work too. Like a lot of things that maybe in some constellation can be automated, but we very much did manually. There's something that they just-- there's no shortcuts to them. And AI in many cases, like we were saying in the beginning, can feel almost like cheating or a shortcut. In some cases it's not there. But that's maybe for another episode. I think number one, obviously the most-- I think no one is too surprised. I'm still thinking, by the way, I don't want to interrupt you. But I'm still thinking in my own brain, like talking with myself when I'm using chat GPT or Gemini, whatever for content. And I'm using it massively. I'm admitting here I'm using it all the time. It's another tool. It makes me better and faster. But I'm still arguing with myself if I should replace the M dashes. If I want to hide it or I don't give a shit, I think that today to hide the fact that you are using AI, it's also very fun. Yes. And that's a very interesting debate. I actually spoke to someone recently that told me that he used to use a lot of M dashes. And he doesn't write with M dashes anymore. Chris Thief has said, why are you setting stuff that's generated by AI? So there's no winning here. Not with the M dashes. I think for content writing, obviously, it helps. I mean, it just makes everything faster. Not necessarily to come up with ideas, but more so just dumping a bunch of information and speeding out the blog. Great. We've been doing this for a very long time. But something that's been important for me from the beginning is there needs to be a lot of human involvement in everything we do, like some of the people listening to this might be surprised to hear that half of our LinkedIn posts are not AI generated at all. Like sometimes I just sit and write them. And that's like what they are. And I think with AI, there's also something where we've become very scared to just be natural, or maybe make mistakes, or something is enthrased appropriately, because if you can write it better, then why wouldn't you? And fair enough. But on the other hand, everybody else-- Like maybe the standard went higher. The standard went higher. But on the other hand, if people are in consuming this content, because they know that it's AI generated so they're bored, so you're like, you just keep through like a lot of these posts. Like when I see things on LinkedIn that I just know, like, come on, there's people that kind of put two sentences together for the love of God. They couldn't write two sentences properly together. It's interesting. It's interesting. Because of course, we solve the company, and at least reputation-wise, it's very cool. And people are coming to us, especially like early-stage entrepreneurs. They want to speak with me and to get maybe consulting and a few tips and mentoring. And when some of them are sending to me emails or WhatsApps of two or three sentences, and it's full of mistakes, both typos and grammar stuff. And I think that today, my level of acceptance to that decreased dramatically. There is no room for those kind of stupid mistakes with the tools we have. At least I think so. Maybe I'm too aggressive here. But the impact of AI on the standards is interesting. But then let me ask it differently. You see someone that you know your familiar ways. How they speak, their style. And suddenly they post something on LinkedIn or any other forum for that matter. That just doesn't sound like them. Doesn't read like them. It just reads like a standard AI response. So my opinion in practice, it won't work. Like let's take our example. We've been posting from my profile for a couple of years, almost every day, even before prom. It went pretty successfully. And people were trying to do it as well. And they thought that the secret sauce is essentially just to commit to that and to post every day. And now with AI, you can do it very easily. But when you have nothing interesting or unique to say, and you don't have your own language and music and character injected into that, and you just copy paste some AI-generated content, I believe that it won't work. People feel, if it's bad, like the type of mistakes example I gave on one end, one side of the spectrum. And they feel also the exact other side. When it's completely AI-generated with no character into that, both of them are not legitimate today, in my opinion. Yeah. And it gets even better, like not to digress too much from the conversation. But it gets even better when people post something on LinkedIn. And then with the LinkedIn premium feature, you have the AI-generated responses. So then it's just AI talking to AI. And the content is just so bad. There's nothing authentic of it. And everything just looks exactly the same. So we can also talk about, like, we're not to use AI. But when do we use AI at prom? So yes, we always use it for content generation, for summaries, for, like, yes, all things written text. But there's always human involvement. And also, there's a lot of knowledge that we've gathered on notion, on Slack, on Google Workspace. Like, we're not just like putting content together that doesn't have any foundation that's based on what we're doing. So I think also the work of all the teams of, like, making sure that everything is documented internally has helped so much with just being able to create content that's external facing. In terms of brand alignment, very early days are marketing agency that worked with us on the website. And like, come up with this concept. And it's still visible, because we still do it, for the thumbnails of the blog. But like, well, we're an security company. It's OK that it's obvious that our thumbnails on the blog are AI generated. And we have this prompt that we craft that I'm with edited a bit with the evolution of the brand. But we have these tons of years for image and video generation to make sure that everything stays no system, follows certain style. And then we just insert whatever variable there is about the topic you want to exemplify through that visual. So it has definitely helped with that. Back in the day, it wasn't thinkable, right? We would have to wait for like two or three minutes. day turn around from a graphic designer to be able to send you something to post dates. In a now-run hub, you can have a blog post with visuals included, ready in a matter of minutes. But I think that's a more conventional thing. I don't think I'm innovating or sharing anything too new here. There's also all the part of data analysis. I love playing with HubSpot and creating reports. But at some point I just need to turn it around with the fact that I didn't have to export the CSV and run a lot of my different formulas if I wanted to know what's the title of our buyer person. I could just ask a question. And that's also open a brand new world of possibilities or we're consuming data and how we're analyzing it. And that's with features that are built in the CRM. I didn't even have to create anything new there. Also for idea visualization, you know I have a lot of ideas. Some of them are bad. Some of them are good. Some of them are better if not to execute. And some of them are okay. But it's always easy to explain it to a human because humans have that much patience. Our mascot, for example, at Black Hat. Some of the listeners might have seen that there was an inflatable 10-row triangle of our logo strolling around the Mandelae Bay during Black Hat this year. I see for that. I'm missing it because we literally signed the M&A I think the same day of exposing the brackets. So I'm fully waiting for this year mascot. Yeah. No, definitely. I mean, I would say the mascot was high up there in events that happened that week. But yeah, I think that that position definitely tops it. But I remember thinking I want this. But like, how do I explain that I want a vendor of triangle that's inflatable? But how does it go? I knew it would take a lot of iterations to do this with a graphic designer. So I just like played with Chad G.P.T. back and forth, generate an image. And then we passed this to a 3D printing or like a mascot vendor in the US and printed it. And like that's so many other things. And now, fast forward to 2025, there's tools like Lovable that I never in my life thought that I would be building things on like that I would be building M2N a website or a mini tool. Back in the day, committing to building a mini tool for go-to-market purposes especially, that was like a gigantic thing. It meant you needed time from product, time from R&D, you needed to show clearer why. And this could be like a three to four months project with unclear signs of whether it's going to be successful or not. In a matter of a couple of hours and some prompting and maybe some HTML edits that I think any marketer can do, like you can put something out together if it works great, double down on it, otherwise you sunset it and it's fine that the stakes are fairly low.