Using AI to Multiply Your Marketing Without Hiring a Bigger Team

0:00Hello everybody. I'm so excited to be here. Well, we are so excited to be here. Um, yeah, today talking about how solo founders and nimble teams can leverage AI to multiply their marketing efforts. Many founders think a little bit like this about marketing, like if it could be fully automatic. If I could just focus on building my product, I would be super happy. Somebody else was like, "Ah, it's not that hard once you learn it." And I agree, Mark already told you marketing is not that hard. But sometimes you just don't want to do. You don't have time for it. You don't like it. And this is also one that I really

0:38enjoy where they say building is like falling in love, but marketing is like trying to convince the in-laws. Now, as a marketing nerd, I kind of disagree with that, but I can see why that would be the sentiment because marketing is a lot of things. It's creating original content. posting on LinkedIn, posting on Reddit, answering questions in groups, doing customer interviews, organizing partnerships, doing shorts, going viral, launching on Product Hunt. There's so much that goes into marketing your SAS business. And on top of that, if you are the solo founder or maybe you have one person in your team that's doing marketing, then this just becomes an

1:19endless to-do list. So, let me ask you, who here is solo and is like trying to do their own marketing? All right. Yes, that that's what I was thinking. Who has um marketing team or at least a person that's like responsible for their marketing? All right. Anybody has more than five people in their team? All right. Cool. [snorts] So, when I was CMO of Thrive Themes, we had a team of seven. And even then, it

1:50was still like this endless to-do list. It was still always things that we had like really cool ideas. Yeah, I'm this weird one with the glasses in the front was that we had so many ideas of things that we could do and that we would never get around to. And a lot of those ideas required resources that we just didn't have. So dev resources for building tools for example on the website and that was just never a priority in the marketing team or the marketing for the developers was never a priority and this is yeah why our to-do list even with seven people was always longer than

2:29what we could get to. But that is also why when about a year and a half I started seeing how much AI was improving and I was like oh my god if I was still CMO at that point like it's so cool what you can do now with AI and because you can automate like real marketing tasks and that's what I started to play around with because when you hear about AI like a lot of it it's just like the hype like so Many people are just like, "Oh, automate a thousand blog posts and you can make these videos with yetis that are like drinking beer and it's so

3:08cool." Like, okay, but how does that fit in with my marketing? How does that fit in for my SAS? And I started to test it out and I still was really excited about what was possible. And so I started building these things out and started sharing it with people. And the reason that I was still very excited is because I believe that small teams, solo founders are now able to get 10 times the result with smart systems and with AI fully automated. No, I don't think that's a thing. I don't think that should be a thing. Um, but 10 times faster or way more efficient. Yes, that

3:50is definitely possible. And that's what I will share today. So, I will share four automations that will do just that. It's basically to give you some ideas of where this could fit within your marketing strategy. Again, we've I've seen some SAS companies who've done like cool stuff with video. Um, and I think especially for something like e-commerce, like the image generation is amazing. But I will be talking more

4:18about other processes. But first, let's get one thing straight. I'm not going to talk here about like oneclick of a button magic pill solutions because I'm assuming that you care about your product that you care about your customers that you care about the product that you're solving and about your reputation and so you don't want to put out AI slop that might work today but it's just going to tank your website tomorrow right I'm also not only going to talk about SEO because I know that for a lot of uh of people marketing kind of equals SEO But yeah, to me that's not the case. So, can we agree on

4:58that? I did have fun making the Yeti. [laughter] This is not a banana, by the way. So, first of all, what is AI exceptionally good at? And Rob already brought up some of the things that I completely agree with. So, AI is really, really good at data transformation. It's really good at pattern recognition. And it's really good at doing those boring, repetitive tasks that you just don't want to do, that you don't have

5:30time for. So, first of all, data transformation. What do I mean by that? Well, if you have long form content, if you have a webinar, for example, and you can make it into a blog post, you can make it into a newsletter, you can make it into a LinkedIn post, you can make it into tweets, all of that. AI is really really good at the example of the podcast where we take the transcript of the podcast and automatically we get our our notes and maybe some shorts and already some tweets that we can use. It's also really good at taking messy data and making it into a structured output. So if you have

6:03a call transcript and you want to make it into FAQs for example, that's something that AI is really really good at. And then translating, so translating either literally so from one language into the other. I've heard of uh someone here who said we translated our website into 10 different languages thanks to AI and that was really part of our growth. Um but also translating it from one team to another. So this might be technical documents that then you translate into

6:30support documents for example. The next thing is pattern recognition. So these were some of the examples also of how you can add AI to your product, right? But this is also how you can use AI in your marketing. So one of those would be like finding common threads in support tickets. This is the category where when you are in the thick of it, when you're doing marketing, you know that you should be doing this and it's just never at the top of your list. You just never really get to it. where it's like reviewing those customer uh those customer reviews that you're getting and seeing if there are any common patterns

7:07in there or those call transcripts from when you were doing customer uh discovery calls and going through them and asking questions. Those are things that AI is really really good at. And then finding nuggets in those customer reviews that you could use on your landing pages, that you could use in your social media. And then the boring repetitive tasks. So if you have to do the same thing every week. So maybe it is uh finding 10 people to outreach to. Um that data collection that scraping can be done to AI. Uh tracking competitor data for example is also a really good one that can just be done through AI or preparing

7:49performance reports. Basically any time when you find yourself copy pasting AI can probably do that for you. Now an example of this is also those those tasks that just are not worth for a human to do but that can bring in a lot of value once they are done. So I was working with a um with an NGO and they had a whole bunch of photos for example from their missions and each time these photos just went onto a Google Drive and preferably in a folder that was like

8:21June 2023. So the moment that they needed to find those photos for their marketing efforts, the moment that they needed photos for PR, for their website, for their social media, somebody had to go through those folders and look in June 2023 or maybe it was August 2024 in order to find a picture that would be relevant to what they were talking about. Now, we set up a system where AI would look at the picture, tag it, add a description, and now all of a sudden their Google Drive actually became searchable. So the moment that they needed those pictures, they were available. This this was never going to

8:57happen without AI. A human was not going to do that. Nobody was going to sit there and label those uh those images. But now they became available for the team or for another AI agent to actually find the right social media photos that they could post. So that's kind of the task that I like where it's add like it creates more val more value than there

9:20was before the AI. So let me show you some automations that I think are very exciting. You might not agree on that but so first of all how much time do you spend on building new features versus promoting them? So, I know that often the building of a new feature is really exciting, but then the marketing work that comes next to it where you actually have to tell your customers what that new feature is and what it's doing, it it's a bit less exciting, right? And especially because you feel like if you tweeted it once, everybody knows about it or if you send

9:56an email once, everybody knows about it. And if there's one thing that we know is that that's not true. I have a really interesting story about a launch about that, but you can ask me over coffee. Um, and so this is for example a new feature launch agent. What does it look like? It's just a form where you fill out what your feature name is, uh, what the launch day launch date is, uh, who this feature is for, like some of those basic questions, and then we let the magic happen. So different AI uh not agents it's it's because agents have tools but uh different AI prompts that

10:38are working all simultaneously and all of the sudden you get your launch content back just a couple of minutes we have tweets that are written a blog post that's written a newsletter that's written and this can be like this is can be adapted to be completely in the tone of voice of your company to make sure that it fits and you can go back to building. So the result all your launch content created in seconds you can go back to building and you know that at least the the effort that you put into building is now compounding because people are also getting notified about

11:17what you are doing. All right second automation how are you showing up in LLMs? Now like I said I'm not going to talk specifically about GEO. There's a talk from Jesse right after this who will be talking you about strategies. But it is interesting just as a starting point to know what do LLMs actually pick up when they visit your website. What do they see? Who do they think your product is for? What do they know about your offerings? And you probably want to test this like several times. Like maybe once a month you want to make sure that you check in. Maybe you adapt some copy on

11:58your website based on the information that you're getting here. And so this is an LLM company analysis agent. I really need to work on my names by the way. These agents not very sexy names. Um but yeah, so in this case you would enter your company name and URL and then the AI browses the web, figures out what it thinks about you and you read and compare the results. So this agent is a browser agent and so it actually goes through using different LLM. So in this case it's Gemini, Chad GPT um and Claude uh and we have those three a those three LLMs look at your website and then you

12:40get back a report where it tells you what your product does or what it thinks your product does, who the customer is, what the price point is, but it can also flag certain things that are not on your website. And so it's and it's a super simple prompt to be clear. It's literally just like hey what do you know

13:00about my company or about this company. We don't say my company about this company. And then it will just also flag to be like oh but I couldn't find any testimonials for example or in the FAQ this wasn't very clear or um it says that it offers this premium value but we couldn't find any uh case studies that would um um say that that is true to actually confirm that claim. So getting this information back can help you just make your content better on your

13:32website. The third agent, YouTube. So, YouTube is great for discovering software tools. And Laura was talking about our uh partners and why it is so good to basically get somebody else who already has an audience and already talking to your audience to talk about your product. And YouTube is great for that, but it's a lot of work to find those creators, to reach out to those creators, to try to negotiate an affiliate partnership, maybe an a content partnership with them. So, what if an AI tool could do that for you? Could find those micro influencers that you can then reach out

14:16to. So, in this case, um our that looks a bit weird. uh we fill out your company name, your company URL, and then you describe a little bit about the company. So what's the product exactly? What's your target audience? And then the agent go to work. So this is both an AI agent that will find keywords that um are related to people who would publish about your content on YouTube. So for example, if you are a form solution, then it would look at things like Jot form versus Google forms for example because if somebody already published videos about your competitors, then they are much more likely to also publish videos about

15:04yourself or we're looking at things like the target market. So somebody who's already talking to um small entrep to entrepreneurs or to SAS businesses or whatever, they are more likely to then be inclined to talk about your product too. So then we have some some scraping going on. Um and in the end you get a huge list with creators with information about the channel and you can go through it and be like yes or no. This would be a good fit for my company. I want to

15:33reach out to these people. And it just saves so much time between having to do that manually, having to go to YouTube, getting lost in a bunch of cat videos, and one hour later being like, "Oh, I only have three creators on my Excel list and I just lost two hours of my life." Number four, same thing. Reddit, one of those rabbit holes. And if you followed a little bit about what LLMs use uh to find information, then you know that Reddit is one of those important websites where they like to get information about software, about tools. Um, and having those creators and that authority that comes from people posting

16:27on Reddit about your product, then you know that it could be really valuable potentially to answer on posts on Reddit. But again, if you have to do that manually, if you have to find those posts manually, it takes time and the risk the risk that you get lost before you even get to comment is really really

16:47big. So why not let an agent do it for you? You enter the keywords you want to monitor. You have AI analyze the posts to see if they are relevant for you. And then you can just receive the links to the best posts so that you can engage with them every day in your email for

17:05example or maybe you prefer Slack. So this is looks something like that. The automation we look for the different keywords. We then look when was the posting date. Um have an AI analyze it to see like okay does this actually match my company? Is there a way that I could engage with this post? And then it sends an email with the uh with the results. So and then all you have to do is visit that link and decide whether or not you

17:37want to engage with the post. So the advantage again times saving and having the opportunity to have something it's kind of boring you know like finding those posts every day and engaging and and you know that compounding work. So if you can do this every day it would probably add up you or your marketing person. Um but you might just simply not have the time to scour through all those posts on your

18:05own. Now, you might be thinking like, okay, all of these like seem like they might help a little bit, but it's still pretty far from, you know, automating my whole marketing. And it's because in my mind, any of these automations, any of these agents, they are just the beginning. Your content can be p published automatically. That launch content, for example, once you go through it and you're happy with it, you could have another automation starting that then

18:33actually schedules it and publishes it. Same with the outreach to the YouTube creators. You have that list, but I still would want a human in the loop to make sure that the people you reach out to are actually people that make sense to your company. Um, but maybe you can just take a box so that then it finds the email address of those creators, maybe their social media profiles, and then it can send an outreach email to those people. Or you can have Reddit responses that are crafted already based on the an on the post, the original post, and then you can go in and be

19:09like, okay, this this is something that I would say or this is not something that I would say. And again, this can be fairly sophisticated because you can feed it all your blog posts, all your videos, and so it can make really, really helpful posts because I think the moment that I'm talking about, okay, AI can create Reddit responses for you, most people are like, oh no, cringe, not a robot, please no. But if that robot has your whole catalog of YouTube videos, has your whole catalog of blog post, can use your voice, can be trained on that and then be like, okay, this is an actual helpful response to this

19:44person that I would would have written or even that this AI can write better than I do because I don't remember which podcast episode I talked about this thing. Um, it can become really really powerful. But where do you start? There's so many opportunities. So where do we actually start? Well, I think that you should build your AI team one by one. You pick one process that you like to get rid of, that you like to get off your plate and then you see if you can automate part of that process and once you do, you add something new. So when you see maybe uh I don't know if your LinkedIn looks

20:25anything like mine but I get inundated with uh AI agent content and there's these these posts of like oh my marketing team is 40 different agents and the truth is yeah it it might be like it it can actually be really really helpful but you probably didn't start out that way like you start out with one process you make sure it's solid and then you build on that. So, just to be clear, I've been showing you N8N automations. Um, I think some of you are familiar with the platform. It's basically like the Zapier but more sophisticated and for uh they were they were very early in the AI game. Um, but

21:12it could also be a custom GPT that you built and then instead of having to write those three newsletters that Mark was talking about on your own, you have a custom GPT that can help you with with writing those. Um, that's just a good way to start before fully automating it. So, that's my pitch for AI in marketing. Um, if you go to the address there, you can test the Reddit agent. You can also test the YouTube agent. Um, there's no opt-in, but if you want to talk to me, you also find a button there to have a little call. Um, I don't know how long I'll leave this up

21:58because [laughter] this is using my API codes for the LLMs and uh and for the scraping. Um, but yeah, that's my vision on AI in marketing. It really is that amplification, getting rid of the grunt work, making sure it keeps your voice so that it stays authentic, that we don't just inundate the internet with AI slob. Thank you. [applause]