Chris Norton: 0:04

Welcome back to Embracing Marketing Mistakes, the podcast all about the biggest screw-ups the world's best marketers have ever made and what they've learned from them. So you don't go doing the same thing. I'm your host, chris Norton, and today I'm excited to introduce Jono Matusky, founder and CEO of Newsprint, ai and Automation Aficionado and a fearful user of Google Ads. Jono has had a varied career. He's also the first father and son combo that we've ever had on the podcast. After having his dad, he launched a not-for-profit incubator program in Cuba until international geopolitics got in the way. He's navigated the highs and lows of marketing campaigns, including one very memorable Google Ads screw-up. He created a brilliant media monitoring tool, combining the wonders of AI and the time struggles of marketers to ultimately create efficiency for marketing professionals.

Chris Norton: 0:53

This is interesting, and in this episode we'll be asking Jono all about his career, his love of AI and agents and, of course, the biggest mistakes that have brought him to where he is today, which you can all learn from. So, as always, sit back, relax and let's hear more about how you can use AI to track your business's performance. Hi Jono, welcome to the show. Hey, thanks so much. Thanks for having me. Welcome as the first son of the father and son combo to ever appear on this show as well.

Jono Matusky: 1:26

It's great.

Will Ockenden: 1:26

Yeah, it's a shame we couldn't be on together we'll do that later in the year, I think yeah, do you ever do shows?

Chris Norton: 1:33

do you ever do interviews with him?

Jono Matusky: 1:35

no, I haven't. Uh, that actually sounds like a lot of fun. My dad's actually in davos at the moment. He's sending us uh photos on the family group chat for their, like, ai summit, which is pretty exciting.

Chris Norton: 1:48

But no, I haven't done any podcast appearances together yet Okay, Well, I mean, when we look at your career varied is the word that I would open with. So do you want to tell us how you got started and what you're experienced in? Because we'll get into the world of AI later on, but I think the fact where you started and how you've got to where you'd be would be quite interesting.

Jono Matusky: 2:10

Yeah, absolutely. So I went to Carnegie Mellon University and got an engineering degree and then quickly got my MBA while I was there and started working for a venture-backed startup that spun out of Carnegie Mellon that was doing energy storage. So we were building large-scale battery systems for renewable energy projects and it was a new battery chemistry called sodium ion that nowadays is getting a little bit more attention. Obviously, lithium ion batteries are the popular one, but kind of came in on the business side. So I was working as a product manager for Aquion Energy was the name of the company and you know kind of got my hands dirty with in a startup which was really exciting and kind of caught the bug and so worked there for a few years and then ultimately left to go help start up an incubator program in Havana, cuba. So this was an opportunity. That was, you know, essentially working with early stage founders and kind of playing you know chief of staff role for different companies are kind of helping them figure out how to launch their services or products abroad, how to build relationships with US companies. So this was in the period where, you know we were US was sort of normalizing relations with Cuba. So this was in the period where, you know, us was sort of normalizing relations with Cuba and there was an opportunity for that and the hope was really to build up, you know, a community and connections with entrepreneurs and ultimately, you know, we ultimately fund some of them. We helped a lot of startups, which was really great. But unfortunately, you know, relations kind of soured between the two countries somewhat unexpectedly, so we had to, you know, essentially shut our things. The hope was that maybe we could reopen them after a couple of years, but it sort of just proved infeasible. So after that I pretty much continued doing the same thing I was doing down there, but just in the US. So started working with early stage you know VC backed companies, again kind of providing just general kind of either chief of staff or more technology implementation services. So started working with companies who you know had some funding and needed some help. You know rolling out, you know standardizing their processes or rolling out new software. And this was sort of where I got more involved with AI.

Jono Matusky: 4:31

While I was at Carnegie Mellon, you know I took material, I took machine learning classes. It was, you know they called it AI. We weren't really calling it AI at the time, but obviously, you know, conversations around open AI and generative AI were starting to bubble up, and that was around the same time that my dad's firm, gregory FCA, got a little bit of private equity funding and so they were looking to acquire some other companies and, you know, roll out new software and also starting they were starting to dabble in the AI world. And so I started kind of consulting for them and kind of helping them with some of the stuff they were working on, and it quickly became apparent that they needed a little bit more than than part-time support. So I kind of came on as, I guess quote unquote, you know chief innovation officer to sort of help you know kind of standard thing, standardized things they were doing across the company, roll out new software and then really build out what we were doing in AI. And so, you know, my dad and I were really early to that to some extent, or at least bringing it over to the marketing side, I mean, I remember when GPT-3 launched that was pre-ChatGPT, and you could sort of generate like we generated a press release using some of those tools and it was really exciting to sort of just see that appear on your screen.

Jono Matusky: 5:54

And then, obviously, when ChatGPT rolled out, probably about a year later. That's when things really took off, but yeah, so started playing around more on the AI side there and then left Gregory FCA about a year and a half ago to start Newsprint, which is my current endeavor, which is we're essentially doing AI curated and AI written executive briefings, so executive news briefings for busy executives and communications professionals. So at Gregory FCA, a lot of the companies that we worked with were very keen on having some sort of brief in their inbox every day that was really focused on the news that they cared about as a company, right? So like mentions of their company, their competitors, their industry, regulatory updates, just any critical information that they might need to know, and in most cases, a human being has to write that, and they got to go read all these headlines and articles and sift through them and collate them and write up a summary, and so we're essentially training AI agents to do that for you as sort of our initial product, and the vision is really to make this something that's available to all companies.

Jono Matusky: 7:02

So any company could you know you could have an internal news briefing focused on the information that's most relevant to you and then also using that as a way to build out. You know analysis of the news and recommend. You know sort of actionable recommendations that PR professionals or communications professionals could use to decide. You know where they should be spending their time or their effort and who they should be pitching to, what types of podcasts you know they should be going on. Things like that, um.

Chris Norton: 7:31

So yeah, I know that was a bit long-winded, but a little bit of a winding journey I know I almost interrupted you halfway through because you mentioned the cuba thing, because the ai thing's fascinating and we're going to get into it. But I was, just like you, worked in Cuba. I went on holiday to Cuba and I think once I came back with a new sense of what communism is all about. And I did because we went into the local town and everything. We took a taxi and saw how the locals lived. A lot of people go on holiday there and don't see any of it, do they? They just go and stay in the five-star hotels. But we went to the local towns and everything and it was fascinating. How the heck did you run a company out there? Was that completely unusual?

Jono Matusky: 8:47

Yeah, so we basically had a US-based nonprofit called the Innovadores Foundation and we were partnering with a non-government NGO down there called the Ludwig Foundation. So yeah, I mean Cuba is a very interesting place, like I'm happy to go into it. But you know, generally, you know think about communism right, and in many cases, you know in many ways Cuba is, you know, probably the closest thing to a true communist state. You know, in terms of compared to, you know where many countries are at now who are sort of part of the.

Jono Matusky: 9:21

Soviet Soviet bloc. But it, you know, you have the state. I mean for a long time that any, any enterprise exists in the state. So you don't have nonprofits, you don't have companies. Obviously, right, If you want to buy something, you go to a state run store and all those things in that state run store are made by state run enterprises. But also, you know, the thing that was sort of surprising is, again, they don't really have non, there's no nonprofit entities, there's no, there's no entities that exist outside of the state really. And so that was so we but there's like a handful there's maybe 12 in Cuba that are technically NGOs make money outside of working for the government, unless you were in very strict categories of entrepreneurship, Right, and so for a long time that was kind of artists who could sell their artwork, musicians who could like travel abroad or who could record albums and sell things like that, and there's a couple other categories you could go abroad and make money, so anyway, so this NGO was sort of a natural partner for us, and so we were trying to essentially build an incubator on the top floor of their building that would be dedicated to incubator on, like, the top floor of their building that would be dedicated to entrepreneurs and kind of using our sort of the arts as our initial entry point.

Jono Matusky: 10:58

So we worked with, like you know, clothing designers called clandestina. You know they, they do these really awesome like shirts and bags, and they sell them. They sell them largely in their store in Havana and then we were helping them essentially launch their brand internationally and, you know, getting them into e-commerce and building partnerships with you know, labels outside of Cuba. You know, worked with video game developers, you know. So, like initially kind of worked with folks who were on that line between artists and entrepreneur and then use that as a way to then start working with more traditional entrepreneurs that you know you might think of in the us, like software developers and folks building apps and websites.

Jono Matusky: 11:40

Um, but it was you know it was, yeah, very, very, uh, educational, as you know, as it sounds like for you as well, you know, when you visit there, especially if you go behind the scenes and kind of see how it really works it's an extremely eye-opening experience.

Chris Norton: 11:57

The state runs and owns everything, right down to the hotel shop. We had a hotel shop and the state run that. The state owns all the traffic that looks modern. Everything else that's local is all old school. They haven't got enough money to do it. And then one of the fascinating things that we had was we had a barman in our hotel who was a brain surgeon but made more money behind the bar because of the tips, because they all get paid the same salary. Yeah, it was so to run a business in a in a purely um communist society like that.

Chris Norton: 12:31

It's fascinating that you were behind the scenes and doing that. It's really, really interesting. I never got to go to Havana, so that is supposed to be an amazing city, so I bet that was a fascinating experience for a good, good career grounder.

Jono Matusky: 12:43

Yeah, it's. It is a remarkable career grounder, good place to start your career, or whatever.

Jono Matusky: 12:49

Yeah, it was certainly. You know you learn a lot very quickly, yeah, and you know I have a lot of respect for, you know, cuba and the Cuban people and the folks that we got to work with. You know it was just sort of unfortunate. I mean, hey, you know I did fine, you know, got to go on and work on other things in my career. But obviously a lot of the folks that we worked with, you know they things didn't exactly go the way that they hoped with. You know they things didn't exactly go the way that they hoped.

Jono Matusky: 13:15

You know many of them are still in Cuba. Most of them have since left. You know they've gone abroad and they're either starting their companies outside of the country or they have gone to work for other companies or they've gone to grad school. So you know it was a really it's tough, you know you tough. You come in as sort of an outsider, as an American. Cuba, um, and in many ways I think like geopolitical forces certainly meant that we had to stop, but it didn't necessarily impact my career particularly negatively.

Will Ockenden: 13:41

But you know, certainly the folks that we worked with, yeah, have had a much harder time kind of navigating that experience um so um going from focusing on um newsprint for a moment, because there's a few themes here which I think would be quite interesting to unpack. So, first of all, you mentioned, while at Gregory FCA, a lot of the work you, you did or your teams did. You know, it's a sort of fairly traditional senior consultant type of role, isn't it? Brief executive briefings, recommendations off those briefings, in terms of what that means from a comms perspective, I mean as a starting point, that's hugely time consuming, isn't it? I mean, have have you got any kind of insight into how long teams will? You know? If you think about an annual contract, I mean, that's a huge amount of time, isn't it every week, in terms of researching what's happening, making recommendations yeah, yeah, no, I we.

Jono Matusky: 14:35

So right now, I think, you know, as as a founder, you're constantly learning and I think, through our interactions and advisors, I think the one thing that we've been migrating towards and now really kind of going down is focusing on enterprise customers. So larger scale companies, right. Who can spend thousands of dollars a month on a subscription, instead of hundreds or tens right. And obviously, newsprint the idea of having a daily news briefing is something that appeals to a lot of people. You know like, we have a lot of individuals who sign up and want to try newsprint. But the reason we're focused on enterprise in many ways is because some of them are spending $10,000 a month, you know, $20,000 a month, thousands and thousands of dollars a month just to get these sort of executive briefings. Because they want, you know, every source that's mentioning forever chemicals, right. Or you know, they want to know about every time their company is mentioned in the news and how they were mentioned right. Or they have 200 different you know energy projects around the country or the world and they want to know about any time one of those is mentioned right. And yeah, that's really time consuming, you know, I mean that takes someone out, you know, or a team, you know, hours a day. To kind of put that together depending on the size of the company, right, publicly traded company that maybe has a bunch of projects or has subsidiaries, you know that. You're talking hours a day. You're talking 10,000 plus dollars a month contract.

Jono Matusky: 16:01

Smaller companies obviously have less news, typically, unless you're a startup, right, and maybe you're getting a lot of attention. So they have, you know, less of a need. Maybe their competitors are getting less news. You know they don. They have, you know, less of a need. Maybe their competitors are getting less news. You know they don't have as many products, right. So there's definitely like a sliding scale of how much time someone would have to spend. But, yeah, I mean, it's a lot, it's a lot and you're looking through a combination of, you know, traditional media, but then also you're looking through industry specific newsletters or members only, uh, you know, uh, news publications, um, just trying to pull all that together and so, yeah, it's an extremely time consuming and, in many cases, not very rewarding process for human being to do yeah, I was gonna say his work.

Will Ockenden: 16:49

Yeah, and and when you're kind of coming up with a concept for an ai tool that solves a problem, do you have a kind of a certain criteria? Because it sounds when you're kind of coming up with a concept for an AI tool that solves a problem, do you have a kind of a certain criteria? Because it sounds like you're looking at those tasks that are time intensive and there's no particular alternative. You know, because if you look at that, it's immediately time intensive, isn't it as a task? So do you look at those kind of tasks and think that you know there's potential to automate that. You know there must be a way to automate that through AI. I mean, talk us through your thought process, I suppose, when it comes to coming up with a solution like Newsprint.

Jono Matusky: 17:24

Yeah, so obviously at Gregory FCA I got to spend a lot of time solving a variety of different problems. So we built an AI press release generator, basically, so you could go in and we had a very guided process. We answered some questions and it wrote, you know, a much higher quality release than you would just get if you just dropped in a chat GPT and tried to get it to write something for you. You know, and we experimented with building out like a spokesperson database, so having like an internal database of all of our clients and you know, and their spokespeople or their, you know, executives who could comment on certain issues and a bunch of other things, right. So we experimented with sort of a bunch of bespoke and just different software solutions to solve different problems internally. And, yeah, I think the reason that newsprint is where it is is that, you know, I saw a problem that companies were paying for right, like whether they have someone on staff who's dedicating hours out of their day to solve a problem or they're paying an outside agency or whatever the case may be, if they're dedicating dollars to it and you can say, hey, we can replace that with AI in many cases for higher quality right, we can write briefs faster than a human being can do. We can write that. We can actually pull in more sources. We can summarize a bunch of different sources all at once and reference them, do things that would be difficult for a human being to do. So that was sort of you know criteria.

Jono Matusky: 19:05

Number one is like what are the problems that companies are paying for? And in many cases, like Gregory, fca actually wouldn't do this work because it's so tedious and doesn't deliver that much value that you know clients would expect. Hey, we want to know about when our company's mentioned, obviously like that sort of PR basics, like they want to know when our company's mentioned, how are they mentioned? Right, they want to get an email about it and they want it to show up in their monthly you know analytics, you know report at the end of the month. But many of them would ask for these sort of news briefings and it largely was something that Gregory of CA would stay away from and it's just yeah, it's. It's just very tedious and the value add is pretty low the amount of time that their team would have to spend doing it versus like how much you could add to the retainer. It didn't really make sense, and so many companies either do it themselves or they, you know, shoehorn their PR firm and get them to do it for them, or they pay a different monitoring, a dedicated monitoring agency that does these briefings, and so that was sort of criteria.

Jono Matusky: 20:10

Number one is like is there a problem that people are paying for a solution for? So that was sort of criteria. Number one is like is there a problem that people are paying for a solution for? And then criteria number two is kind of like what we got out of? Do people really hate doing it? You know like, do people not enjoy doing it themselves or paying the agency to do it? Are they not satisfied with what the solution looks like? And that was kind of, you know, the sort of guiding principle, I guess, or the the reason that you know newsprint is sort of taking the form that it is um, yes, because it it's solving a problem that people feel enough that they're paying for and that they don't particularly enjoy doing and how is it better or different to existing solutions, so, for example, a social media monitoring solution google alerts automatic alerts when a category?

Will Ockenden: 20:55

Yeah, google Alerts as well when a category or brand names mentioned. I understand there is a big difference, but it'd be good to hear it from you.

Jono Matusky: 21:02

Yeah, no, totally yeah. Yeah, Obviously, there's a ton of tools you can use. If you, if you say, hey, I want to know about X, y and Z, you could set up a Google Alert. You could. If you're using tools like Cision or Muckrack, right, you could set up alerts through that. There's a bunch of different tools that offer, you know there's like News Whip. If you want to do social media, there's all sorts of tools you can kind of do to, to, to monitor and get summaries Right.

Jono Matusky: 21:28

Where that usually falls down is typically it it's often a fire hose, right? So you're often getting just a ton of information that isn't particularly useful, and so you know where that often comes out is. If you have a company that gets mentioned a lot, right, because they're publicly traded or they're just a hot company, right, if you try to set up an alert, it's constantly getting mentioned. Some of those mentions matter. Most of them don't, you know they're from publications that you don't know about, or they're random social media posts, and so part of it is just a filtering problem, right? And so how can you use AI to understand, right, what the audience cares about, what the user wants to see, and go beyond just keyword filters, right, so you know. You could say you know, I, you know, just basically, tell our AI, I only want to see news from top tier publications that mention our company in relation to X, y and Z right Mention this particular product that our company offers. Or, you know, I want to know about this competitor, but only in this market, right, you might have a competitor that serves a bunch of different markets and you only care about one of their products, right, and so you know part of it is just using, essentially you know, ai agents, right, so AI kind of working in the background to read news before it ever reaches your desk and kind of filtering it out before it ever reaches your desk, and kind of filtering it out. So that's.

Jono Matusky: 22:55

One big difference is that most other companies still rely on either, like Boolean, traditional searches or they have some level of machine learning, but it's mostly just entity recognition. So you're identifying the name of a company and then pulling that in. So we go beyond that and actually have AI agents that understand what you care about and why. And then the other side it's how do you sort of synthesize or combine all that information from these different sources? Right, so you know it's one thing to say I want a summary of a news article, but I mean, what most folks care about is how was the company, how was our company, talked about, right? Or or what are multiple different publications all saying about this event or this particular issue? And so, you know, are there different viewpoints of how this is sort of being talked about, right? And traditionally that was done with things like sentiment, right, which is very, which is still pretty subjective.

Jono Matusky: 23:52

What you really want to know is just like, how did this reporter talk about this topic, right? Or how do they talk about our company? Like what's, what's a one sentence summary? Or did they, did they quote one of our executives and what was that quote? Right? Did they quote a competitor? You know what are all, what are all the regulatory news? You know what were all the regulatory updates today, just summarized as a bulleted list, and so you know that's sort of. The other half is, once we have that information, our AI can decide how best to summarize it in a way that is as efficient for delivering that information as possible, right? So it's not just a list of headlines, it's not just, you know, full text articles that somebody on your team has to read through, it can actually write. You know a fairly interesting newsletter that just delivers the information that matters to you.

Chris Norton: 24:43

Can it have personality then? So, for instance, could we have a? Because our show is quite. We've been called irreverent because of the mistakes issue and we always talk about marketing mistakes. Could it? Could your ai agent create a newsletter? That would send us a newsletter all about people talking about marketing mistakes around the world at any point? Is that what it could do, and it could write it in an entertaining way. Yeah, like talking about examples that would make good people on the show. Is that?

Jono Matusky: 25:10

yeah, yeah, so, so, yes, um, I mean there couple, couple, couple answers to that in terms of, like the different, different elements of it. Right, but on the one end, could it write irreverently Like absolutely Right, and we've all messed with chat, gpt and tried to get it to you know, write in certain styles or whatever. So we, you know, for our summaries, you can specify, you know, sort of each for the overall newsletter as well as you know what specific sections you would like it to include. Right, and you can completely change that tone, right. And so you could say, you know, write it as if you are an intelligent business analyst. You could say write it, you know, and you can say write it, you know, be funny, you know, focus on the mistakes that people made. So you can tell it not just what you're interested in, but how you want that information summarized and formatted and what kind of style you're going for. It also has information about you as an audience. So, as it's both filtering and writing, it's writing it to the audience, right? You?

Jono Matusky: 26:09

know, it's all part of all great marketing, right? So you could sort of explain what the purpose of this newsletter is, right? So you know it's all part of all great marketing, right? So you could sort of explain what the purpose of this newsletter is. Right? So you know it's it's whether it's for you internally and it's just for you guys to find opportunities of who to have on the podcast, or this is a companion piece to your podcast and you're distributing it to your you know followers, then you know it would, it would understand that, right, so it can write to the specific audience. But then the other side is sort of finding news.

Jono Matusky: 26:38

So right now, newsprint is really focused on sort of traditional media. We're expanding into social media, we're experimenting with podcasts and pulling in information from those sources, but at the moment we are, sorry, focused on traditional media and so something would have to get written up, right? So if folks were writing a thought you know it's like a thought leadership piece in, you know, forbes or you know something like that or they talked about a mistake that they made, that's where we can pick something like that up. They made. That's where we could pick something like that up, but certainly there are other ways that we could pull it in right. So it's one thing hey, they write an article about. Hey, we made a mistake.

Jono Matusky: 27:22

The other nice thing about Newsprint, though, is it could find an article where someone, maybe offhand, mentions a mistake that they made right or it's, or they talk about something, but they don't necessarily use those, those words, and you can sort of say to it I want a newsletter that helps us find, you know, folks who made marketing mistakes, and it will go out and do its best to kind of fill in that newsletter. Even if it's not, it can't just use search terms, and it can't. You know, somebody didn't just write an article that's specifically focused on that, and so that's sort of. Another nuance of newsprint is it can kind of surface information that would be very difficult to find otherwise and that you probably wouldn't ever see.

Will Ockenden: 28:04

Is it?

Chris Norton: 28:05

daily.

Jono Matusky: 28:05

Yeah, yes, yeah, yeah, yeah. So it goes yeah.

Will Ockenden: 28:27

Typically. I mean, we have folks who are like asked about doing weekly, but typically most of our customers want it daily because that's that means they have, they really have the need and they're willing to sort of pay the price we're hoping to get. But obviously the role of the PR is to kind of analyze the media landscape on a daily basis and then draw on their years of experience and make recommendations into what this means for said company. Now, did you say the tool to a degree does that makes recommendations on the back of the insight it pulls in um? And if so, is there any risk of ai bias, example being tesla? Let's say we're analyzing tesla, huge narratives about elon musk being evil and aligning with trump and doing questionable salutes on stage. So how do you kind of mitigate that?

Jono Matusky: 29:08

yeah, so thanks for bringing that up will. Um no, no, it's a really good question, so that I'll say I mean you guys are in the industry right, so you kind of get it so there's.

Jono Matusky: 29:19

so most, most folks who read the news read the news for the news. They read it for the information contained, contained in the article, right, and so that's kind of a newsprint is still focused on right now, which is essentially just what is what is information coming in? Focused on right now, which is essentially just what is information coming in in real time that matters to my company. And so we look back, so we try to not include in the newsletter something that's already been talked about unless there's new information related to that right, Unless there's like a meaningful update. But right now we don't necessarily do sort of trend analysis or like meta analysis of how is Tesla being talked about across all publications, right, and that gets much more at PR needs. Is that sort of meta intelligence, right? So it's not just you're not reading the news because you care about what this specific article or a couple of articles say. You're reading the news because you want to know the sort of meta information of what how has this story been changing over time across every publication that's written about it? Or or who are the journalists that are covering this particular piece of information positively or negatively, or how is Tesla, right, being talked about? What are the different facets of Tesla news coverage, right? And so that is something that we're building out now and trying to experiment a bit more with, and that's something that's going to be more about giving PR and communications professionals tools to analyze themselves Basically, so kind of giving them access to the information that we now have kind of in our database and in our AI to let them ask interesting questions and determine it, answer that question across as many different articles as talk as speak about a topic as possible, and then have it write up sort of a meta-analysis.

Jono Matusky: 31:20

So, like you know, imagine you could, you, you would want to know, um, you know, uh, when reviewers, you know, or when, when people, uh, talk about our online, do they talk about this feature positively or negatively, right? Or what do they say about this feature, right? And then you could basically have the AI read all the articles that ever mentioned that product, kind of put it into different categories, or answer your question and then make that now actionable, analytical, data-driven intelligence, and so that's where we're hoping to go with this, right, that's an incredible market research, incredibly powerful isn't it for for for comms.

Will Ockenden: 32:01

You know, comms professionals, it's. It just allows you to be so responsive and have your finger on the pulse, exactly, you know, really understand the media sentiment at all times. I mean, what, what a feature that would be.

Jono Matusky: 32:16

Yeah, that's the hope. So we're going to start experimenting with that in the next couple of months. It's just one of those things where now we've kind of built out the pipeline and, you know, the data pipeline. We have a lot of information in our database, but we also have a pretty robust AI agent platform on the back end. That, you know, is first and foremost just for generating the newsletters. But you know, you can kind of, you can kind of see it, you can imagine, right, if you're saying, hey, I care about this topic, but I care about this element of the topic and I want it to go in this section of my newsletter. You know, maybe you want that every day, but maybe you actually only care about that right now. You know, because something happened and so that's going to be what we're going to, you know, trying to do over the next few months is, can we just essentially turn over our you know tools to folks to kind of use in real time, rather than just a daily newsletter?

Will Ockenden: 33:04

I'm I'm curious what's next Now? Clearly you don't want to reveal the next product you're going to launch, but you know it's clear between yourself and your father you're AI experts in the world of comms. You know you're not just talking about it, you're actually delivering products. So you talked about the press release tool that Greg built. You know you've got the newsprint tool. What other sort of clear use cases can you see for AI in PR agencies or marketing agencies? You know those time intensive tasks that just make no sense for humans to do them.

Jono Matusky: 33:41

Right, yeah, no, absolutely. I think there's going to be a lot. I think it's going to be an exciting couple of years and, to be honest, if AI didn't advance at all from today, right, and we just had the capabilities we had, we would still see a lot of change just in terms of the adoption and spread of the existing tech and people actually using it as well as they could. And so I think there's a lot of places where this can be used and should be used, you know. So, like write release, our press release generator.

Jono Matusky: 34:15

That's a great example of just like day one AI. That's like hey, I've got this long thing, I gotta write and I want an AI to help me do it. And so, you know, I think a lot of that content generation is going to get turned over to AIs, but I do think a lot the question is, who's going to be reading those press releases, right? Like, is it going to be a human or is it going to be AIs reading those press releases and using that information to make decisions, right, and so that type of stuff. I think we're going to see more and more where, now you have AI, you've content that's being generated or information that's being published so that it can be consumed by AIs right, and I think that will, in many cases, change the way that journalism works right AI news editors, then is that what you're talking about Essentially?

Jono Matusky: 35:04

yeah, I mean I do think there's a future, right. I mean, if you talk about journalists, right, there are many great journalists who aren't particularly great writers and there are many great journalists who are great writers. But there are many journalists who are really good at turning up interesting, important information or have connections in certain industries or have really interesting insights, and in many cases, their job is actually just it's surfacing that, it's muckraking, it's surfacing that information and making it so that it can see the light of day, and writing the article is somewhat of secondary importance. And I think that we will certainly see that where you have essentially AI journalists as researchers turning up valuable information and getting that information then really being summarized by AI or even just being used by AI. So I think that's going to be exciting.

Jono Matusky: 35:54

I think for marketers, what I'm really excited to see is where sort of AI agents go.

Jono Matusky: 36:01

So AI agents being AI that runs in the background, so you don't have to sit there and ask questions of a chat bot and have it, you know, spit information out.

Jono Matusky: 36:08

But essentially, you could ask, you know, an AI agent to say, hey, can you go, you know, write up or do an analysis of who all the key stakeholders are in this industry in this country or something like that, right, and it would then just go out and do you know online research or pull from different databases or use your in-house tools and have access to those and just pull that information together and write up a little report with sources you know, kind of answering that question. And so you know, I think that's we're still at the early days of that, especially in marketing. I mean, I don't really know of any AI agents that are designed for marketers. Most of them are still more so on the like software development side, but I think that's's we're going to see more and more of that. So so you won't necessarily have to sit in front of chat gpt and ask it to do stuff for you. I think soon we'll be able to, you know, ask it to do a task chat.

Chris Norton: 37:02

Gpt, just last week I think, launched tasks, didn't it? So tasks are the agency you're talking about where?

Chris Norton: 37:08

you ask it to do a task and it'll go off and do it. I quite like the idea that you can ask it to go off and create, like look at something or read something for you, and the thing that's annoying about the various different ais and I know you've built platform on it is that you have to stay in the same chat if you want it to remember the the things that you've discussed before. But you're talking about, I like. I like the personalized gpts where you can create them all around an audience, all around a particular frame of reference, and then you can have chats with that particular one. You don't have to keep telling it all the same things over and over. Surely they're going to get cleverer and cleverer, aren't they?

Jono Matusky: 37:46

Yeah, and I mean so. Openai launched tasks and right now their task is super simple. It's basically you can just schedule a chat so you say, at this time of day I want you to answer this prompt. So it's kind of like scheduled prompts that send you an alert. But I think that they're doing it because it's going, I mean one. I think it's because they want you to turn on desktop notifications so that, like when you're chatting, you know they can send you notifications, that, like when you're chatting, you know they can send you notifications. But I think what their plan is to have that be the inroad to more sophisticated AI agents that can kind of do work for you on the background and to your point, of these sort of you know, gpts right, where you can have customized chatbots basically with maybe particular personalities or access to certain information or access to tools outside of chat GPT. I think we're going to see more of that.

Jono Matusky: 38:40

I think where it's kind of exciting is there's a lot of folks experimenting with essentially using different types of AI agents to accomplish different tasks and we do a little bit of this at newsprint.

Jono Matusky: 38:54

So essentially giving certain types of agents certain amounts of information or access to certain tools and then having almost like a router AI agent that can kind of decide based on what you asked. Okay, who should I refer this over to, right? So right now they kind of call it tool calling, which is like I could ask it, hey, I need to do, you know, I need to find stakeholders in this country, right? And it would then break it down into steps and then it could say, ok, well, I know, I have access to this database, or I have access to Google search and I can do a Google search. But you know, now what we're starting to see is, you know you ask that question and there's an AI agent and it says, well, I have access to these six other AI agents. Each of one can do these different tasks. Who should I kick this over to? Like?

Chris Norton: 39:38

a community of AIs, so you have these clusters.

Jono Matusky: 39:44

Yeah, little AI companies. Yeah, that's wild no, it's, but the funny thing is that they all talk in just English, right, like they just talk in plain language, so you can basically watch the chat transcripts from what these guys are doing and what they're pulling together. But, yeah, we're going to see more of that. I mean, there's going to be essentially these little AI companies, you know, and I mean, to be perfectly honest, we're doing a lot of that at Newsprint. We're not certainly not, you know, maybe as cutting edge as as some of the, you know, vc back companies that are really focused on this stuff. But uh, it's going to become a lot more common, uh, and I'm kind of excited to see where it goes so back to the agency world, and they'll probably be creative directors.

Will Ockenden: 40:25

Listening to this, um hating what I'm about to say. Is anything sacred. So creativity, you know the ultimate kind of human-focused task, isn't it? You know, being struck with inspiration in the middle of the night, coming up with an amazing campaign that just kind of cuts through the noise and emotionally connects with people. Are we ever going to have a situation where AI can come up with award-winning creative campaign ideas that connect emotionally with audiences, or is that always going to be something humans do?

Jono Matusky: 41:01

So yeah, so my answer to that is that there's kind of two types of problems to solve in general in the world, right, one is just like it's just a problem you know and you're trying to, and it could be a big problem, like you're trying to solve cancer, right. Or it could be like you're trying to get your car to start right, and those are problems where in solving it the problem doesn't get harder, like your car doesn't try to stop you from starting it right. Cancer, I guess, and diseases like viral diseases can actually develop, you know defenses and make it harder for you to solve the problem as you solve it. It's like a moving target. Right In marketing, in the business world, many of our problems are different in that as you get better at marketing, so does everyone else, right, and so if you come up, if it gets easier for everyone to make these brilliant campaigns, you're all competing for the same eyeballs, right, and the same attention. So it's an arms race to some extent, right. So it's who is better enough that you can break through and kind of solve that problem and be the one who comes up with something truly innovative, right, and so everyone's going to start. Everybody already has access to ChatGPT, right, and so everyone's going to start. Everybody already has access to chat GPT, and so I think it's going to increase the quality, but we're still working. Everybody's competing with everybody else, and so I do think there's always going to be room for that human innovation, because you're going to.

Jono Matusky: 42:26

I do think that humans will still be able to come up with the thing that the AI didn't Like.

Jono Matusky: 42:32

The AI is always going to be deficient at something, and there may be different types of AIs Some are good at this and some are good at that but there's always going to be gaps, right, and human beings have always been very good at kind of filling those gaps, and so I don't think that that, hey, I have this brilliant idea for a piece of marketing, this ad that I'm going to run. I don't think that's going away. I just think it's certainly going to change, and the hope is that you'll be able to do a lot more, much more creative stuff, much more quickly with less money using some of these AI tools. But it's certainly going to be an interesting time to see what happens and how much better. Hopefully, every ad that we see on tv is is brilliant and compelling and you know, funny or emotionally riveting, um, and you know, I think ai is going to make that easier, but it still means that you got it back in the day, though, back in the 80s, because I'm old, uh.

Chris Norton: 43:31

But back in back in the 80s, tv uh, advertisements in the UK I'm talking here because obviously you're in the US, but they were a lot better quality or seemed to be. People remember things like we had various adverts in the UK and I've heard various podcasts talking about this, but it had its heyday to TV. We had three channels at one point, obviously we had four chat. We had three channels. At one point, then we had four channels in the uk. When we got channel five in the uk, five channels it was like a revelation. So people would watch the adverts. But whereas now not so much so, and when I've listened to creatives that have talked about crafting adverts, they've said that often, like ai is logical, it form, it follows logical um patterns that you're looking at and and putting, it's a computer at the end of the day, whereas creativity isn't logical sometimes. Sometimes things that would never have made sense, actually the things that stand out the most. I'm trying to think of a great example, of something that was well, the um, the classic is um, the cadbury.

Will Ockenden: 44:35

I don't know if you had it in the states, the cadbury's ad of the gorilla playing the drums to phil collins, um, and it was like a two minute ad of this gorilla playing the drums to a classic phil collins tune, it was one of their most successful ever? Ads, yeah, but what logic would possibly say you know what our audience love?

Will Ockenden: 44:52

gorillas, they love the drums they love. Phil collins, let's do it, do you know? I mean, it's just, it's. It's completely illogical, isn't it? And there's a, there's a ton of ads. I mean it's reassuring actually to hear you say that there's always going to be a role for the, the, you know, the, the and the, the human touch when it comes to creativity, and I agree with that. Actually, I think, I think you're right. It's just a suite of tools, and that's also what I find quite interesting. Um, the quality of ideas gets better, so that in itself presents a problem. It's then much harder to cut through. So let's say, we increase the, if we increase the quality by 100, great, everyone's got amazing ads, but then everything becomes beige because no one can cut through.

Jono Matusky: 45:31

Right, right, right. And so I think that if you can cut through, whether it's because you're really good at using the tools or because you do have these really innovative ideas that people really pay attention to, I mean, a great example. So, david Lynch, you know, the film director who I'm a big fan of, just passed away, unfortunately last week, and you know he made Mulholland Drive and the TV show Twin Peaks and his stuff was extremely weird, you know, very outside the norm, but it really resonated with people and he won Oscars but was also sort of this like critical darling and a lot of people loved. I mean, you know tons of people watch Twin Peaks and you know he's an example of someone who nothing that he did was logical, right, you know, and he couldn't really explain a lot of what it was just sort of, you know, just sort of this intuition around what he should be putting on the screen and you're, you know, I think, yeah, in the age of AI, you're going to have a lot of this.

Jono Matusky: 46:24

You're going to have Marvel movies, right, and you'll have smaller scale Marvel movies. That'll be sort of, you know, maybe CGI will be replaced by AI generated video, right, and it'll just be logical right, like what is a superhero story? Who's the superhero today? You know it's kind of serialized, but that does leave a lot of room for, I think, true, yeah, creativity, innovation and doing things that no one else is doing, and people who do that can get a lot of attention, whether it's, yeah, the Cadbury ad or whether it's, you know, david Lynch or whoever. I think that there's a lot of room for that and I think there always will be. So, like, I'm really optimistic about AI, but you know I'm also in the space, if I wasn't optimistic about it, I don't think I'd be room for it.

Chris Norton: 47:12

We are all about mistakes on this show, and you've sent one through to us about Google Ads. Now, google Ads is for someone who deals with data and engineering. Google Ads sounds like the logical step in a career for a young John O Matusky. So what happened there then? A young John O Matusky?

Jono Matusky: 47:30

So what happened there then? Yeah, so this is an example, I guess, of maybe thinking that you're smarter than you are right, but I so, post Cuba, like I mentioned, I was working with a bunch of startups, kind of helping with chief of staff, process development, software implementation and things like that. I generally don't have a ton of official experience in marketing specifically per se. You know, obviously as a founder you have to learn a lot about marketing, but the types of stuff that we were implementing wasn't necessarily marketing software. But we had a client who was looking to. They had an agency they were paying to basically run their Google ads. And they asked us, you know, they felt they were paying too much for the agency. That's always a red flag to somebody comes for you because they're like, hey, can you do what they were doing, but for cheaper, you know? So they asked us that and it seemed like pretty straightforward, right, for the most part. They just needed, you know, to keep things running. So we took it on and at one point they asked us essentially to duplicate one of the ads that they had, but, I believe, just like change the target audience essentially or change some of the keywords that they were targeting. And so we did that. You know, we thought successfully, but it turned out that they were essentially using.

Jono Matusky: 48:47

What we didn't realize is that the previous agency had set up mirroring in the Google ads, so they had terms in there where the ad copy would reflect back what I mean. You guys were talking to an audience of marketers, so I feel like I should be the one explaining this. I'm sure everybody already knows. Well, what I didn't realize at the time was essentially that we knew this right.

Jono Matusky: 49:08

We knew that essentially if somebody searched for this was for a nutrition company. They searched for, you know, nutritionists and we in our keywords had added certain competitor mentions. So if somebody searched for a competitor of our company, they would see our ad. But it also meant that we could potentially reflect back the name of that competitor in our ad, right? So if they looked for, like you know, noom help page, we would. We would in the ad say you know, click here for Noom and not click here you can, but like Noom help page, right? So people were clicking on our clients ad thinking because it said literally the name of the competitor they had searched for, and I guess enough of those got reported to Google that their account got flagged, which is like an extremely rookie mistake, right?

Chris Norton: 50:10

How much did the account spend?

Jono Matusky: 50:16

extremely rookie mistake, right, how much did the account spend? So, yeah, a fair amount. I don't want to say how much. I mean the money wasn't totally wasted, in that many of the ads didn't feature, you know, ads mentioning competitors, right. So, like they, they still got a lot of. You know, that was a thing is we didn't see it in the analytics, like we were still getting the click-throughs and like things seemed totally fine.

Jono Matusky: 50:31

But at one point somebody did contact, send us a contact thing, and was like hey, you know, I'm looking for. You know, is this, is this Noom? Like I thought this was a Noom thing and somebody like flagged that to us and we're like, yeah, that they must've just gotten confused. They said they came in through Google ads and we're like they must've gotten confused, and so that was a stressful couple of months. It was like figuring out how to deal with that. But the client was fairly understanding, you know we gave them a little bit of a discount on our services and said that we would, you know, work pro bono and get things back up and running, and we did manage to get them back on Google Ads after a couple of months.

Jono Matusky: 51:10

But it was just, you know, one of those experiences where you know there's certain cases and this kind of goes back to the AI thing, right Like there's certain cases where you should just hire an expert, right, and maybe that expert will be an AI, maybe it's, you know, a human being. But there's certain things, like Google ads, where the cost of messing it up is a little bit too high and unless you want to become an expert in Google Ads because it is fairly complicated and there is a lot of nuance and it is this arms race, right, you're competing against everyone else who's running Google Ads, many of whom are very, very well educated in Google Ads. And so, unless you have a particularly creative approach, right, or you have a message that you know is going to resonate, or you have a very specific audience segment, that's like kind of easy to reach. In many cases, it just makes sense to hire somebody to run it for you, right, and rather than trying to bring it in house or having your, you know, consultant doesn't really know what they're doing try to do it for you, but, yeah, so I think that's something that I've definitely internalized and have brought to newsprint of kind of know when.

Jono Matusky: 52:15

You know kind of make, buy or rent right, Know when it's worth hiring somebody to do something versus, you know, buying an existing piece of software that can just do it for you, versus paying an agency or something like that, or paying an outside expert that can just get things set up for you and make sure everything's running right. Yeah, that was a particularly embarrassing mistake because it just really seemed like, uh, you know, total work to move, like it's like the worst possible thing you could do. Like if somebody asks you to run google ads, it's one thing. If you run the ads and they get no clicks, then you know that's bad.

Jono Matusky: 52:45

But if you run the ads and you actively get them, uh, deplatformed uh, yeah, that's literally worse than getting no clicks isn't it exactly right and it's a new client as well, a new client, the first, the first project yeah, exactly, exactly, yeah, we just turned off, yeah, exactly, I saved him a couple months of marketing spend, you know so that was um.

Will Ockenden: 53:03

That was absolutely fascinating, johnno. I mean I think, yeah, we've, we've covered ai um before on the show, haven't we, chris? But I think that's a really interesting deep dive, take on a real sort of specific area of it, and it's fascinating actually as well to hear you talk through your kind of thought process in terms of identifying what that need is before you develop a tool. So that was great to hear as well.

Jono Matusky: 53:25

Thanks, no, it was great talking to you guys.