The Exchanges

Q Uh, how do you describe what the company does? Cause on LinkedIn you say custom AI and automation agency. I don't know if that really does it.

A You know, I've gone back and forth on the description of it many times. Most of the time when I'm talking, you know, more informally in person, I just say I run an AI engineering firm, um, which I think is, you know, I used to say consultancy, and people then thought that we were mostly doing advising work, um, and strategy work, which is actually, like, less than 10% of our business. Like, a majority of our, almost 80% of my team are engineers. And so I, I typically I'm like, we're an AI engineering firm. We specialize in generative AI apps. And, uh, if you, all the stuff that you read about on LinkedIn and Twitter, as it pertains to LLM's agents and stuff, we built that for businesses.

AI assessment note: “I just say I run an AI engineering firm... We specialize in generative AI apps”

Q And you've literally done this, or this is a hypothetical?

A I've literally done it. No, I've literally done it. I can show you my Stripe account. People are paying for infographics right now. If you go to strategy.evaai.com, you could see examples of the infographics. Strategy.evaai.com slash infographics. You can see the examples of the infographics right now, and if you guys want to buy it, there's a checkout link right there. That entire, that entire landing page was created by AI. I didn't create that landing page at all. I don't use people to create landing pages anymore. The landing page, the Stripe link, the examples, everything was done with AI. I used to have to use people to create this and to make landing pages and to, no, no, no, none of it. I go to Claude, I say design it for me, I take that design that's broken. I give it to Gemini. I said, it's a beautiful design. It's not working. Fix it. Gemini fixes it. I copy and paste it into go high level. I press save, and the landing page is live.

AI assessment note: “I've literally done it. No, I've literally done it. I can show you my Stripe”

Q All right. So that's no, number one is do an AI service for people. We talked about the headshot. Number two, make a course using AI. We see how to do that. Number three, you told me before we got started, offer a product or service that takes no time to create. You gave me the example of infographics. How are you doing it? Who are you selling it to?

A It's very, very simple. If you want, you can, you can pick an industry. For example, I have picked an industry called the restoration industry. So Ajax Union is going to make a million dollars in profit in 2026 from offering things to the restoration industry. Do you know what the restoration industry is? Water damage, fire, after fire, whatever. So just for example, I'm going to pick one franchise called ServPro. And we created infographics for ServPro to use with their clients to explain their process of how it freaking works, of how this stuff works. And so I said, I have infographics that are ready done, available for you. It's a hundred dollars. If you want the infographics with your logo on it, I will hand it to you for a hundred dollars. If you're not happy with it, I'll give you your money back. People are paying a hundred dollars for infographics that banana, the nano banana pro created. And they're thrilled because they have an infographic they could use. Now I'll give you 10 infographics for your business for a hundred dollars right now, specific for your business. Now, could you imagine you have a collection of infographics that you made in advance, and you have all the prompts, and you go to a business, and you say, within 24 hours, I'm going to work very, very hard, create these custom prompts for your business, and make these incredible AI-generated infographics,…

AI assessment note: “I'm going to pick one franchise called ServPro. And we created infographics for ServPro”

Q Okay, and so then you put that into, uh, where, where is it? I'm on Evie. There's a profile.

A If you go to Save Profiles, yeah, if you go to Save Profiles, you can upload bulk. There's a place to upload bulk. Or you can go to, Or you can go to LinkedIn and press save profile. When you go to their profile, if you have the Chrome extension, you can press save profile. It'll download everything from their LinkedIn into heavy AI, and then it'll give you the ability to find them on Facebook, Instagram, Google, Twitter, X, whatever, and then it'll give you AI agents that sit on the side that you push one button, and it researches everything about them all over. It'll tell you what podcast they've been on. It'll tell you what articles they've written. It'll literally tell you everything about them, And you can write notes about them as well, and you can put them into lists, and it'll show you the recent posts that you can comment on if you want.

AI assessment note: “If you go to Save Profiles, you can upload bulk.”

Q blowing my mind with the stuff that you were doing and everyone else around the table felt like that. But you said, I am learning it and people are paying me to learn from me. And as a result, I've got pressure to learn. I've got an audience and I've got more reach because of this. Is that still possible, though, or do we have enough AI teachers out there?

A Dude, you have no idea. Tony Robbins just came out with an AI course. You know when Tony Robbins comes out with an AI course, and then I join his AI course, and he's not teaching anything that's super valuable in there, and people pay a thousand dollars to learn how to create a custom GPT that, by the way, sucks, so I'm in his community, and, and I'm just adding value, and people are signing up to my products and services Just because I'm adding value, and there's 600,000 people that attended his class. 70,000 people decided to sign up for a dollar. 50,000 people are still paying him 37 dollars a month. To do what? To be in a circle community that he didn't even create the whole thing. It's just a circle community. It's just insane. Look up the AI Advantage. Look them up on school. He has a school community that's free. He has a circle community that's 37 dollars a month. I'm in it, and all the people are in it are hungry to learn AI, and they're not really teaching AI. They're teaching you how to create a very basic custom GPT, and you know what? Those are all potential customers for all of us, for you, for me, for anyone that wants to teach AI. And there are hundreds of communities like that where they're not really teaching anything. And if you can actually add some value, if you could teach someone how to use notebook LM, do you know what notebook LM is?

AI assessment note: “Those are all potential customers for all of us... for anyone that wants to teach AI.”

Q You didn't have a human being code up this site for you, even a human being with AI? You did it yourself this way, the first version?

A I built MVPs. I have about 200 additions to Evie AI that I already built that are working that I created myself using artificial intelligence, and when the refactor's done at the end of this year, in another two weeks my refactor will be done, the developer will go through each one and slowly, modually add them to Evie AI. So I do have full-time development team working on Evie AI now, Um, that is doing things correctly with a proper back end, with a proper integration, with proper safety, with proper compliance, um, with proper privacy, with proper cybersecurity rules. All that's being done by a human being, but the initial MVP, the initial minimal viable product of something that has a front end and a back end, I do that myself. I'll sit down, brainstorm with AI, because I'm the product manager, really. You know, like, I'm the ideator. I'm the person that knows what the real issue is that the customer needs to solve. And if I had to work with a developer back in the days to work with a guy named Andre, I would give him the idea, he would come back to me three weeks later with an MVP. Now, I give the idea to Gemini, and within 45 seconds I have my MVP done.

AI assessment note: “the initial minimal viable product of something that has a front end and a back end, I do that myself”

Q All right. So these are basic, uh, software examples. Heavy is a lot more sophisticated than that. Heavy essentially helps. It's like a, a LinkedIn manager designed to help people sell and to create content, right?

A That's an AI. Yeah. It's an AI assistant for sales for social media. So for example, if you want to write content, every AI will write up all your posts from now until forever. It'll create your images. It'll help you draft comments on posts. It'll help you write what messages to send people. It'll save profiles and research to people instantly. Like everyone in the room that was in the room when I spoke to a group of agencies that you were talking to about bootstrap growing, um, I knew everybody that was in the room and I knew everything about them because EVAI put together a detailed list of every single person and who their target market is and what their messaging is and what they've achieved and all their contact information and everything just for me knowing who's in the room. I told EVAI. It did its research. It put everything together. So it uses agents.

AI assessment note: “Yeah. It's an AI assistant for sales for social media.”

Q all the, all the likes and the hearts and the attention. We're clearly not there. And if anything, it's actually sending people over to Zapier, but it's, it's a nice way to introduce this new tool and tell your followers about it. Now let's see the, the comparable experience. Like if you and I were going to build this type of, uh, brief in Zapier, what would that look like?

A So here we've got, uh, the Zapier agent builder. Okay. Now this is a little different. So instead of, you know, having the like nodes and drag and drop, which you can do with Zapier's workflow builder with the Zapier agent, you're just describing what do you want the agent to do in plain language? And the nice thing is, um, I know a lot of folks struggle with thinking through like all the detailed instructions that agents really benefit from. One of the things you can do with the agent is just describe it Like, you would be talking to a, a, a friend. And you wouldn't give a friend, like, all the nitty gritty details. And so you might just say, uh, make me a meeting brief agent that, you know, gives me some details before my meetings. So this is a pretty bad prompt. Like, this is not what agents generally prefer to get. They like detailed instructions. This is not detailed instructions at all, but we're gonna start with anyway and see what happens.

AI assessment note: “So here we've got, uh, the Zapier agent builder.”

Q Do you have any personal ones that you do for yourself?

A I'll tell you, so the, the one I think is, I get a lot of value out of is, We use, um, meeting recorders for a lot of our meetings internally. And so, um, I have found that AI is an, is an exceptionally good coach. It is infinitely patient. It is neutral. It just observes what it sees and it tells you what it, it thinks it has observed. And so for our exec team meetings, we, we have these recorded. And so we'll take the transcript automatically. We'll send it to a prompt. That is based on the five dysfunctions of a team. So if you haven't read five dysfunctions of a team, it's, it basically describes like good behaviors and bad behaviors of teams. And then based on this prompt, it will generate coaching for all the meeting participants. And so they'll all get a DM that says, here's how the meeting went for you. And it is so valuable because again, it has that just very unbiased, very neutral observer. You know, it doesn't care about, oh, Wade's the boss. And so I need to butter him up a little bit, or, you know, This, I don't really like this person, so I really want to, like, you know, grind their, grind their gears a little bit more than the X person, or like, this person I like a lot, so I'm just gonna be, you know, I'm just gonna take it easy. It doesn't, it doesn't think, it doesn't care about any of that. It's just, it's just like, I saw that here you interrupted this per…

AI assessment note: “the one I think is, I get a lot of value out of is”

Q I've seen whole agencies built on nothing but zaps. They take on a client, they create a zap, and the zap now is getting smarter and smarter. In fact, to call it a single zap is just oversimplifying it. Like what are the smart things that it's doing now because of AI that weren't possible maybe a year ago?

A The biggest new capabilities you get with AI are two things. One, you can add AI into workflows, and what that gives you is the ability to work with unstructured data. Now, unstructured data allows you to do some very specific stuff. One, you know, in the old school world, you might have an email template that was using a mail merge. Well, mail merges are, like, sort of personalized, but not really personalized. I sort of pretend like I know you. But with AI, you can actually go do deep research on somebody, so I can say, hey, tell me about Andrew and Mixergy and everything he's trying to do, and then based on everything you go learn, now I want you to go write a specific email to Andrew using these key points and translated it to him in a way that he cares about. So instead of using those, like, old-school vanilla mail merge, I'm able to actually generate a very custom email that is very specific to everything I know about you. That's something you couldn't have done before. So that's a great example of it. Another great example is You might have, um, a business where you get a lot of customer feedback. Like maybe you've got a bunch of support tickets coming in, or maybe you've got a customer feedback form. Well, in the past, you might want to go look through all that feedback and try and understand, you know, what are the common complaints I have? How do I go generate an FAQ?…

AI assessment note: “The biggest new capabilities you get with AI are two things.”

Q I have no idea you're doing all that. Ok, images too. You and I were talking before we got started about Kling and other tools that you're using because it's not just for text, it's not just for video. What are you doing with images and what AI tools are enabling you to do that?

A Yes. So we use, uh, a bunch of AI tools, all the standard ones, including run, uh, you know, runway for video, mid journey for creative. Uh, what we're seeing very interestingly with this is we're doing a lot of always on creative content. So imagine you go on DoorDash and Instacart's app, you will see these banners floating around, which are offer banners or which are talking about deals and all of that. Yeah. Now imagine, and you'll be surprised, most large enterprises still do it humanly, in a manual manner, including the push notifications, including the creators and all of that. We've created these custom engines, which can produce banner images, add creatives, add skills like 30,000 creatives over the last six months. And we're eB testing this with data to see which creatives work. And then automatically that's queuing an engine to the LLM engine to refresh that creative, to see how we can write more CTRs up. Okay. And with scale at say Instacart or DoorDash level, you'd be surprised that this translates massively into revenue. So we're seeing CRM teams or customer lifecycle marketing teams and all these big consumer apps heavily iterating on, uh, AI usage. In, um, no push notifications, in anything customer comms, which could drive more revenue and acceleration on.

AI assessment note: “mid journey for creative... We've created these custom engines, which can produce banner images”

Q know if something works? They're not telling you what you should be doing, right? And you're trying to get an advantage over anyone else who might have any publicly traded, publicly available information. How do you, Ethan, know if you've done a thing that you got a result because of the thing that you do, and how do you get that iterative process fast enough so you can keep iterating?

A The way that you know anything is true is with a scientific method. So you come up with a hypothesis, you test it and you reproduce it. So how do you come up with hypotheses? You come up with hypotheses by doing analysis. So you do analysis and you say something like best website builder. It seems like these sites are ranking. Maybe that the reason why is for this, like, it seems like YouTube is appearing. So hypothesis, so analysis. Pattern match. YouTube videos are showing up. Red threads are showing up. Hypothesis, adding a YouTube video will, uh, show up in citations and will cause my answer to, will cause me to become an answer and to rank better within the answer. So then you set up a analysis. You take a control and, uh, and a test group. So for Webflow, there's a hundred topics or a hundred questions we want to rank for. For 50 of those, we're not going to do anything. And for 50 of them, we're going to do something. We're going to add YouTube videos for those 50. I'm going to set up tracking for all 100 for, let's say, for two weeks, four weeks, so that I have a baseline, and then for these 50, I'm going to add YouTube videos, and for these other 50, I'm going to do nothing at all, and then I'm going to see if the ones with YouTube videos, uh, showed up. So did the YouTube video appear in the citations, and did my answer go up? And then I look at that. The reason why t…

AI assessment note: “The way that you know anything is true is with a scientific method.”

Q Okay. All right. So help center, uh, what about how to articles on the site that's helpful is, is help center enough?

A The more, the better. An interesting thing when you search for, when you search for, uh, rippling and pave, or sorry, when you search for which payroll management software integrates with pave. There's multiple articles that show up saying rippling. There's a rippling pave integration page. I believe there's a pave rippling integration page on pave.com. And I believe there's a help center article. Another example is a meeting note transcription that works with zoom and otter shows up with a feature, a zoom feature page, a, uh, help center article saying how to do this. And I believe there's a page on zoom, but basically there's multiple pages and all of these are getting otter to show up. And so it's as I wouldn't say that, you know, the more, the better forever, but I would do multiple versions, help center article, how to do X article, paying the company that you're integrating with and ask them to do both of these things.

AI assessment note: “The more, the better.”

Q And so do I now need to create all these different answers? Isn't the AI smart enough to grab three different answers, combine them, and say, here's, here's what my user needs?

A Yes, you do not need to create millions of pages for every single question. You need to generally answer the question. So an example of that would be, Which, which, uh, for Rippling, uh, I want a payroll management software that has these different features like, uh, uh, attendance tracking and, and, um, time tracking and off-boarding and this. So Rippling has a page for each persona, uh, HR, IT, finance, and all of the use cases and features that they have, but there's one page for each. So it's not that you need one page for every single use case in every single Question. It's grouping and theming them. And the AI is indeed smart enough to just find the answer. So you don't need it explicitly verbatim exact match for every single question.

AI assessment note: “you do not need to create millions of pages for every single question.”

Q It's not so much that people aren't using search engines. They're gonna use it for a long time. There's some people using the white and yellow pages, I think. It's not that it's dead that it doesn't work. It's just that there are killers like you out there who are so good at it, I can't compete. What do you say about the SEO is death worry that I have?

A Yeah. So there's a few different, let's break it down into a few different arguments. So argument one is what you just said, which is it's too hard. It's crowded. And I, I mean, I do think that it's hard. Uh, I think it's hard for early stage companies, which is why I mentioned when, when a startup asked me what they should do in SEO, I say, don't work on it and wait until you're larger. Wait until you're, you know, series A or later, maybe series B and later, because it's, it's, it's hard. Like maybe you could get a little bit of impact, but you're not going to get millions of visits if you're a small company. But if you're a big company, you can do really well. It's, you know, it's a massive channel. It's, uh, I think it's the second largest channel after paid advertising. So, uh, I think it's five X, at least five X LMs, maybe 10 X, depending on the category. So it's way bigger than, than AEO. Uh, so it's a massive channel, but yeah, I mean, it is hard. So depending on how big you are, you should decide whether or not it's, it's too hard, but it's not that I am so good. It's that authority is You need enough sufficient authority to compete on competitive terms. But if you have that, then, then you can be competitive. So it's not so crowded that no one can do it. It just depends on who. So that's argument one. Argument two is that, uh, SEO is, or sorry, search is being used l…

AI assessment note: “it's not so crowded that no one can do it. It just depends on who.”

Q How can I tell which questions I should be targeting? Like what are the top asked questions?

A Yeah, there's, there's, what would be great is if ChatGPT gave us a ChatGPT search console, like Google search console. I said, here's the search volume for these prompts, but they don't do that. I don't anticipate they'll do that soon. I think that they'll eventually do that as they build out ads, but they don't, they don't give you that information. So there's a couple of ways to do this. The first way is what I would suggest, which is to use search data. And so if I'm a brand new company, I would start with the keywords that I want to rank for and then make them into questions. How I would pick the keywords would probably be which keywords would I want to pay the most for. So either I'm already doing paid search and I use my paid search keywords, or I look at my competitors paid search keywords. So those are the money terms. So if I'm, uh, you know, if I'm a Rippling competitor, I would go look at Gusto deal in Rippling and, uh, see what, what are they doing for paid search in the top paid search keywords, and then just make them into questions. The other option is to use, uh, Subsets of real data, first party data through browser extensions or ISPs and things like that. Or like, there's a, you know, a bunch of LM companies that might sell you actual real data. Like there's many, many chat bots you could buy for, you know, chat GPT won't sell you their data, but maybe a less…

AI assessment note: “So there's a couple of ways to do this. The first way is”

Q It's the AI assistant for email because it goes through big email inboxes and then tells you what you need to pay attention to, and you can go find the rest in an orderly way, right? That's the business, plus a newsletter people pay for. Okay, um, I've heard a couple of things about you. Number one, you say we're mostly bootstrapped. What does it mean that you're mostly bootstrapped?

A Um, basically, we've taken a little bit of money, so we raised 600 K in 2020, which is when we started the business. Um, and then we just raised another round From Reed Hoffman and, um, starting line DC. And that was a two million dollar round, but, uh, we raised it as what I call a sip seed round. So they have committed up to two million dollars, but we can pull down the money whenever we want in whatever amount we want. And we've not, we've pulled down like 500 K or something like that. Um, so, uh, you know, a little bit over a million dollars raised over the last five and a half years. And done in a way that, um, you know, when I, when I, when we raised money originally, I told all of our investors, you should expect that this is the last money that we ever raise. Um, and I think, uh, the investors, this, in this round also are, are, um, aware that we're probably not going to go raise, you know, ten million, uh, ten million dollar series A from Sequoia or whatever. I don't even know if ten million, maybe ten million is too small now for a series A, but yeah. I'm a thirty million dollar Series A from Sequoia, you know, in 18 months. Um, we really try to run the business at Breakeven. Um, and we're growing really fast. Uh, I feel very ambitious about what, what we, uh, what we are doing. And also it's very important to me to preserve, um, this spirit that we have, that we're a…

AI assessment note: “basically, we've taken a little bit of money, so we raised 600 K”

Q What do you tell them to do? How do you solve that problem for them?

A The first thing is, is they need to understand what actually moves the needle from a business from a revenue standpoint. So you take all the marketing things that you guys are doing. Is it pay-per-click advertising? Is it SEO? Is it GEO like ranking on chat GPT? Is it social media marketing? Is your ads? Is it your emails? The list goes on and on. Once you can figure out what's actually driving revenue, Then you can figure out, alright, what out of that Arsenal that we're using to drive revenue isn't doing well. And what do we think we can improve upon the most? And you have to look at both what isn't doing well and what you think you can improve upon the most, because if something isn't doing well, like tick tock and I'm Boeing, I'm just using a hypothetical here and I'm selling airplanes. Well, no matter how good at tick tock marketing I get, I don't think tick tock is going to help me sell more hundred to two hundred million dollar airplanes, right?

AI assessment note: “The first thing is, is they need to understand what actually moves the needle”

Q on, right? You can imagine a world in which ChatGPT does all this or Google's Gemini and its consumer grade products that people are using to figure out what to do with their, with their weekend plans that it's natural for them to then ask the same question for. It's not an unfamiliar tool. How do you, when you do this, how do you compete with, with these bigger players?

A Yeah, absolutely. I mean, I guess I have two answers. One is obviously the one that, um, we're, you know, at Y Combinator, we, ah, you know, we believe in that we see it happen. Like, we just see small teams of people go out into the world and create, you know, they, they're obviously using these incredible frontier models, but they're adapting them to the very specific things that real people in the economy actually need, and so they're not necessarily glamorous Scenarios. Uh, there are customer support scenarios for, um, HVAC consultants, for instance. This fragmented industry, but a very big industry, um, And they're taking, they're building software, you know, there's a company called Avoca that we work with at YC. They're, they're doing exactly this customer support for HVAC, but V-one of it was basically Service Titan. So Service Titan is a incredible public company, but, um, they're basically software. And, uh, HVAC consultants and firms spend about one percent of their dollar wallet, you know, for every dollar For every hundred dollars they bring in in revenue, they spend about a dollar on software like Service Titan, but they spend five or six dollars on actual people picking up the phone and doing scheduling and doing all that stuff. So the wild thing that we're seeing is that if you, like, scope what you're doing and make the thing that is perfect for that set of peo…

AI assessment note: “they're adapting them to the very specific things that real people in the economy actually need”

Q Who did you think the customer would be for that?

A Initially, I thought it was going to be sales. I think you were, I think the way you've alluded to makes sense. It's like on a sales call, you want to know when it isn't going well, because then you can stop, pivot, and ask a question. You can say, I'm going to take a beat. Do you understand this? We have sales, but what we realized kind of in 2023 was AI is making this mainstream where people are going to find applications that we would have never expected. So one example was someone in 2023 reached out to us really early when we started getting product market fit. It said, Hey, I'd love to jump on a call. I've got a use case and I'd really like to talk to someone and I'm going to have my caregiver join the call as well. We're like, okay, this is interesting. Did not expect this. All right. Jump on the call. We're like, Hey, I was like, Hey, I love your product. This is great. Um, let me tell you how I use it and why I've got the caregiver on my call. Uh, I have early onset dementia. Um, I forget things. Uh, I know I forget things right now and I want to, I talk with my family once a week and I want to remember the things that we talked about so they don't feel bad. And so the caregiver was there to actually say, hey, how do we keep this subscription going while this person is having the conversation where it is going to get worse over time because he's looking at these notes …

AI assessment note: “Initially, I thought it was going to be sales.”

Q be connected to meetings, I, anyway, um, Um, I'm, I'm obviously, I'm very excited about the space. It really is so fun. I'm excited about your software, but let's go back to the original sales thing. Were you thinking that you would sell this to salespeople and they would have that sentiment analysis in real time or for post game analysis to see what they could have done differently? Both.

A So we started as purely real time that then you could look at it after the fact, because if you're a salesperson, you get one opportunity to make the pitch in a lot of cases. So you want that instant feedback. Uh, what's not realistic is your manager sitting on every single call that you have to give you that feedback on Slack, on Teams to say, hey, slow down your pace, talk about this, you forgot about that. Uh, it's not realistically, that's not realistic. So from our perspective, we thought that real time feedback would be incredibly valuable. Um, what we found was It is valuable for that niche use case, but a lot of times it becomes cognitive overload where if you are doing poorly and all of a sudden the score goes down from an engagement standpoint, people are like, oh shit, what's going on? What do I do? How do I solve this problem? And what we found was the feedback was like, this is great. I actually believe this. This is pretty accurate, but you're not telling me what to do next. And so that's where we were kind of like, okay, what do we need to do as a next step? Beyond just showing the metrics went from 80 to 75 to 69 to 62, how do we change that outcome?

AI assessment note: “So we started as purely real time that then you could look at it after”

Q What's your background in AI? How did you know that this was possible?

A I think my background in AI was more machine learning. So early in my career, I worked at a company called Faircast. We did airfare price predictions. So in the old days, this is really old days, you'd actually have to have boxes in different areas to actually go on the internet, pull data down, and you'd build these things called scrapers. And they would go in and say, scrape the airline ticket prices from all these websites on a daily, hourly basis. And we actually built a model that could predict airline ticket prices. And so that was really kind of Early days, machine learning, a little bit of AI, depending on how you want to look at it, to go in and say, we can predict airline prices based on date, time of day, uh, routes, weather, all these different variables that come into play. Uh, then when we started placed, uh, kind of very similar thing. If you remember the early days of the Apple phone, that blue dot would bounce all over the place. It would normally be a giant blue dot that someone, you're somewhere in this neighborhood. You go to Chicago, you'd look like you're on the other side of the river. And so with all those, those noisy data points, Uh, were signals for us. So we actually built models to go in and say, Hey, it says you're across the street, but if I send you a push notification to check into the nearest location that you have, where are you? Oh, you're at…

AI assessment note: “my background in AI was more machine learning. So early in my career, I worked”

Q users are using Read and what else they, what else they would want and what they're not? I guess what they're not using is easy enough. You check the software and you see what people aren't using and you maybe mark it a little more or understand that it, that it hasn't worked, but the things that they need that they haven't expressed, how do you come up with those?

A Uh, part of it is the metrics that you had talked about. I think the other part is kind of going in and just getting feedback. So because we have 50,000 new users every single day that are coming in, they're constantly trying these things out and we add new features. Uh, we use a company that recently was acquired, StatsSig from OpenAI, where they're able to let us test different things out to go in and say, hey, should we focus in on Google Drive Connect or SharePoint or OneDrive Connect because people want to search their meetings and the files that they've created to make sure that we can connect the dots there. And so we do these tests. Stats like lets us do that. We're able to see what the behavior changes are very quickly, and we give a winner, and then we turn that across the board. Or it's feedback where someone says, like, hey, I need this in Portuguese, and this is where we, this, Brazil is our top five market for us. We would have never known that had it not been for the customers kind of just coming out and telling us and saying, like, hey, we need Portuguese, and we're like, okay, are we seeing a big uptick in Portugal? I was like, actually, no, it's Brazilian Portuguese. It's a little bit different, and so we went in and we said, Hey, are we going to build two separate models for Portuguese and Brazilian Portuguese? And we're like, yeah, I don't know. The user sta…

AI assessment note: “part of it is the metrics that you had talked about. I think the other part is”

Q and make sure that they're getting their work done. I would imagine that there would be a note taker for the different role in a company, and I'm not seeing that here. And again, I'm looking at you guys as the model for what's the future of a lot of AI SaaS because you're so far ahead. So why do you think that it hasn't broken down that way here?

A I think you're, you're in that scenario, you're applying traditional SaaS approaches where it's like, I've got a solution for sales. I've got a solution for coaching. I've got a solution for intro meetings. And in this AI world, you don't need that. I think that one AI solution needs to go in and say, how do I tackle this for an entire organization where you're interviewing customers, where you're doing sales calls, where you're having internal standups and having that one solution actually gives you much more value than an individual standalone. So to give you an example, Let's say you've got a sales team of 50, they're going out in the market, they're pitching the product, clients are asking questions. That might be in a solution like Gong today. And you've got Gong, but it's too expensive, like you had mentioned before, where it's like, it might be too cost prohibitive, and the product team doesn't use it. Well, now all of a sudden, the product doesn't have access to any of that information. That is siloed in this wall where it's like, salespeople are talking with customers every single day, telling them why they like the product, or why they don't like the product, or what features that they want. Where we become is the system of record for productivity. And so that's going in and saying, bring in your emails, bring in your messages, bring in your Google Drive, your SharePo…

AI assessment note: “in that scenario, you're applying traditional SaaS approaches... in this AI world, you don't need that.”

Q Give me another example that either you or your, your students have built. What else?

A Yeah, we, uh, did a big series of lessons recently that, you know, essentially is intended to be a Uh, template and a demo that students can then take and use for their clients. And we call it a dashboard for team operations. And the use case that we showed is actually very similar to what you just described the agency having built for your, um, for your friend. Uh, and it was basically we're a museum. We have a bunch of cool stuff in our museum and we know a lot about it, but like, we need to get that knowledge out of our heads. Into a system. And then we want to generate a social media post every day and make sure that we have really interesting content that we're pumping out on a regular basis. So very similar to the idea. So we built that. We actually, what we do is we build three levels. We build the easy version, the medium version and the hard version, and we show the students that progression. So then they can pick and choose where they want to be in terms of their stack. So it can be anything from air table. All the way up to a more complex NN or make build that's going to be, uh, you know, talking to different APIs and doing more complex stuff behind the scenes. In this case, you could bring in like a buffer or a Hootsuite, for example, to, uh, you know, do the social media posting. So you can kind of sprinkle in as much complexity as you want, but the fundamental val…

AI assessment note: “we call it a dashboard for team operations. And the use case that we showed”

Q Is that too much? Is that like for me to ask for a system? It doesn't have to be that system, but it needs to be systemized. Am I being too anal there?

A No, I, I actually have a system I'm going to send to you, and we can link it in the show notes. It's called the decision filter, and the decision filter is a one pager that has you outline what the project looks like at perfect or at complete. It shows the five success criteria. It shows the ROI. It shows whether you're going to make money or make people happier. It's a great way to hand off a project to somebody because they understand why we're doing it and how it should look in the finished state. Almost like a creative brief for a TV ad is going to look. Um, so if you delegate that project, she'll understand it and she's competent enough to pull together. So again, something that's good enough. She'll write an email that's good enough. She'll handle it good enough. You'll see a few of them go and then you can take good enough and make it a little more perfect later.

AI assessment note: “No, I, I actually have a system I'm going to send to you”

Q How do we think of you as an A, like more, more of an A plus student than I am?

A No, not at all. I was always a C. I was 62% in high school, 62% in college, but I realized that no one was ever going to look at my transcript, so it didn't matter. So I became president of my fraternity. I was on the university ski team. I was doing student government. Um, I was running a business. I had 12 employees when I was in second year university. So I was very active and engaged in everything else. And then school was kind of, yeah, I'll get that. And I think if you can bring a bit of that into your company, you win. I was coaching a CEO the other day. He'll do six million in revenue this year out of his dorm room at Cambridge. He's in third year school at Cambridge. Made two million in profit and he's, he, he was wanting to quit. I'm like, you can't quit. If you're at Oxford, Cambridge, Harvard, and Yale, you stay in school, you finish, but I want you to not go after a four point O be okay with getting a 2.8 for the first time in your life. Call mom and dad and say, look, I'm going to get a 70% average and I'm going to make three million in profit in my last year. They'll be happy. And I think for you, it's a bit of the same thing. You, you, if you can move towards momentum, creating momentum, And not perfection creating momentum. You know, no one ever said perfect creates momentum. So just getting microphones out the door to the guests will wow them. It'll impress th…

AI assessment note: “No, not at all. I was always a C. I was 62%”

Q Why? What was it that you were looking for in those calls that allowed you to know this is the guy who I can work with?

A I knew he had the skills. That was actually kind of, um, non-debatable on the skills, but on the culture side, he was such a Renaissance man. He was such a high gloss individual that he'd done so much at 23. I was worried he would piss people off. I was worried that he was going to be so strong that people wouldn't like working with him. And culturally he would be a disconnect. So when I called his boss, I said, I'm calling about Christopher Ian Bennett. And he said, Oh my God, you've got my, my best guy. And I said, you know, I just have one question. Like, do people like working with him? He said, do they like it? They love him. And so then I knew, right? When, when, so, but you have to do the groundwork on, I don't want to wait 90 days to find out if I was right or wrong, because, because there's too much, the ripples that that can cause can be really damaging. So you just, you have to slow to hire quick to fire.

AI assessment note: “I just have one question. Like, do people like working with him?”

Q almost, almost too good to be true. Step out of the way. Let me do that for you. And I could not only trust them with all my banking information, but also let them go and do that. And that's, that's possible. Have you seen that done? First of all, you're nodding and smiling as I say that, like, yeah, of course that's possible, right? I want to acknowledge that.

A Yeah, it's, it's very, it's, it's not easy to do, but it's much easier to do than we think if you really are clear on what you're looking for. And if you spend the time on the front end to make sure that you're hiring the person who has those traits, who, you know, like I had drinks last night with a friend who I hired 16 years ago, and we were laughing about the fact that I did 12 reference checks before I hired him, seven business and five personal. And the final business reference check I did was to his boss, a guy named Mark Hamilton. And I phoned his boss while Christopher was standing right beside me at my desk. Christopher knew that if I called his boss, he'd lose his job, but he also knew if I called his boss, his boss would tell the truth and he would get the job at 100 got junk. And he's like, no one would have ever gone to that depth. And I said, I had to, it was a big decision for me to hire him. And I wanted to do my groundwork. Well, I actually had a VP and a director tell me not to hire him. And I'm like, no, you know what? Everything I've done is checked out. I know you two are wrong. Turns out that those two were gone within six months, and Christopher Bennett went on to become an icon at one-hundred-got-junk.

AI assessment note: “it's not easy to do, but it's much easier to do than we think”

Q I've talked a lot about the reasons for this. Let's talk practical. Give me an example of someone who's actually using ManyChat to reach their customers and how he's doing, you know, as a business, not like, is he having interesting conversations?

A Sure, sure. So, we have this, uh, just messenger alchemist, uh, Sean, uh, Patrick Simpson, who, uh, gets, he, he got over 40,000, uh, messenger subscribers on his messenger list, and, uh, basically what he does is he uses our growth tools, and inside ManyChat there is a section that allows you to, like, convert visitor, block visitors into messenger subscribers, uh, like, links, uh, Facebook ads into Messenger subscribers, and he uses those to start an interaction, and Facebook has this policy of 24 hours after the first interaction, not the first, actually, the last interaction, so every time a person interacts with your bot, you have this 24 hours to basically message them for free and about anything. So it can be promotional, it can be ads, After 24 hours, it cannot be ads, so it's, it's kind of like the policy that Facebook has right now. So he has this brilliant, like, sequences and, uh, interactive messages that he uses to drive traffic back to his, um, store, uh, online store, and to actually do sales. So one of the campaigns that he did, he, um, um, spent, like, 2000 on ads and, uh, made 11,000 in sales in just, like, a few months. Days. So, it was really, like, the ROI that you can get on new marketing channels is just crazy.

AI assessment note: “spent, like, 2000 on ads and, uh, made 11,000 in sales”