Transcript 0:00 We get to build all the cool stuff from sci-fi that I grew up wanting to see in the world. AI can learn everything about their brand, and after onboarding, they could be publishing within 15 minutes. 0:10 That's as far as my knowledge goes, though. [laughs] Carl's probably watching this going, "That's not true." Well, that's not true. Yeah. [laughs] Close. Close. 0:19 Last week, Cuttable announced a $20 million Series A led by AirTree, the same firm that backed A Cloud Guru. It came with a relaunch. 0:27 Cuttable now pitches itself as an AI ad platform that learns, and it works with more than 440 eCommerce brands, from Lululemon to Cotton On. So this week, we're flipping it around a bit, and Sam's our guest. Yeah. 0:39 I get to ask all the questions. [laughs] I'm a bit worried about this. [laughs] It'll be more fun than ever. I can put you in the hot seat. [laughs] Yeah. 0:45 Um, so welcome to Imperfect Company, uh, the podcast where great founders tell the real story about building businesses, and we're gonna put Sam through his paces today and go into- Yeah... 0:54 uh, the real story of building Cuttable and some of the things we don't hear in the press, and go behind the scenes a bit. Yeah. It's been a wild ride, so very happy to chat about it. 1:02 Um, I think in the press you always hear, like, all the good, amazing things, and there are a lot of good, amazing things, but you don't see the craziness that happens behind the scenes to get there. 1:12 So yeah, very happy to chat to it, for sure. So we've talked about A Cloud Guru in the past. Mm-hmm. Maybe very... 1:18 In a minute, the background there again, just to refresh people that haven't heard your, your story before. Yeah. 1:22 I started a company called A Cloud Guru with my brother in 2015, and we were teaching, uh, teaching engineers how to build things in the cloud. So teaching geeks how to be geeks, basically. Um, selling videos online. 1:35 It was an online learning platform. My brother and I started that from our spare rooms in our, in our houses in 2015, just the two of us. 1:42 Um, over six years, scaled to 600 employees, um, over 100 million USD in ARR, and, uh, sold the business for, for $2 billion in 2021. So that was, like, the definition of a wild ride. 1:57 [laughs] Um, the ac- like the biggest challenge there was just scaling the company- Yeah... to match the demand. 2:03 Um, so yeah, that, that's my background, and then took a year off and was crazy enough to start another business, which was Cuttable. And before you get into Cuttable, why go again? 2:13 And I can somewhat relate to this, but I- Uh, to, to be hon- I just like building things and I think... I, I took a year off. I didn't really sit still for that year. 2:20 I started building little apps and I just, I just like building, and I get bored if I'm doing nothing. 2:26 And AI started happening in 2022, um, and I just didn't wanna miss that wave of, like, this is, like, fr- from my perspective, what's happening now is we get to build all the cool stuff from sci-fi that I grew up wanting to see in the world, and I didn't wanna miss being a part of that. 2:42 Yeah. Yeah. Well, I mean, the- Like a-... the, the whole, I mean, the cloud, you know, SaaS kind of wave was, was big, but this is just so much bigger with AI. This is wild. Yeah. 2:52 'Cause we can, we can just do so much in the creative spaces now, into the analog world. There's so much we can touch we couldn't touch before. Yeah. Um, yeah, and it's just fun. Like- It's more fun than ever, yeah. 3:03 Yeah. It's so much fun to build right now. Yeah. Yeah. And so Cuttable, how did that... What was the genesis? Um- Other than wanting to build another business. [laughs] Yeah. I... 3:14 So I had been building an app for my kids, um, like a little storybook app. Oh yeah, I remember that. Yeah. Yeah. Back on the... Uh, so I built an iPad app. 3:22 This was like 2022, 2023, the first voice models and image models were coming out. Mm-hmm. 3:26 So I built, I built this app with my kids where, like, they could talk to an AI character on the screen and it would, like, build them a story, like a visual storybook, and it would narrate them and then they could, like... 3:38 It would do six pages, and then they could say what they wanted next, a bit like a choose your own adventure. Yeah. So I was kind of building that just to show them what's possible. 3:45 Um, and so I got really, like, into how do you merge together voice models, image models, um, uh, LLMs. And then at the same time, I met my two co-founders, Jack and Ed, and, uh, Jack was from the advertising world. 3:59 Run, runs one of Australia's best boutique advertising agencies, Sunday Gravy. 4:03 And, um, yeah, they were just seeing this pattern where customers again and again and again needed more creative, advertising creative for their- Yeah... social media ads. 4:13 So think of, like, you know, video ads on Instagram. To feed that algorithm, you need to feed it a lot, like a high volume of n- new ads, new ads, new ads all the time. It's c- it's called creative- Yeah... 4:24 but really it means an ad. Yeah. Yeah. Um, and so you go from a world where the advertising industry's set up to make, you know, one or two ads a week if you're lucky. Yeah. 4:34 At, at, you know, like three ads would cost you 30 grand to get made, or it would take weeks, to where brands need to be publishing 50, 100, 200 ads a week. Yeah. Um, and so you have to do it a different way. 4:46 And so when I saw this problem, I realized, hey, actually, all this stuff I've been doing with image m- models, voice models, LLMs- Yeah... this is creative production. Yeah. 4:57 This is the first time that computers can solve a high-volume creative production problem. Yeah. 5:02 Um, and so there was this intersection of this massive industry need and this new enabling technology- To solve the problem... to solve the problem. Yeah. Um, so yeah, I thought, "This is the thing to jump into." Yeah. 5:14 And so I ditched the kids app [laughs] which was fun. I loved it. I pulled it off the App Store, um, and jumped into this. Yeah. Yeah. So then you raised a pre-seed or a seed, got some capital in, got going. 5:27 Yeah, raised a seed with Square Peg. I think that was 2023 or 2024. Um, and yeah, we, we... So we spent six months, frankly, building the wrong product. Yeah, so we'll get into this. So there's a- Yeah... 5:41 essentially you had to pivot the company after that, right? So you're six months in. Yeah. Raise your seed, you get going, and what were you building that... 5:48 What, walk through, so what you were building and why change it? We were building for... We had all these connections with enterprises through my co-founders and- Yeah... 5:56 relationships, and so we just had these customers ready for, "Hey," like, "Hey, we need more automation. We need more high-volume creative." 6:03 And we realized, though, what we were doing is we were kind of- By, by building for enterprises, we were building very custom solutions. 6:11 You know, enterprises, we're, we're trying to fit into a workflow of, you know, 50 different people and, and- Yeah... a bunch of agencies. 6:18 And so they, and they all had a different setup, and there wasn't really a lot common- Yeah... that, that we could engineer for. 6:24 And so we found while we had this automation technology, there was still a lot of human involvement. Mm. 6:30 Um, and it actually culminated in this famous Slack message inside the company where our head of engineering, Kyle, posts a zip file, "Here is customer X's Black Friday campaign." Mm. 6:42 And he's posted it in Slack, and we were like, "What are we doing? Our head of engineering is literally generating a specific customer's ads [laughs] you know- Right, yeah... himself." It wasn't scalable? 6:53 Was that the- It wasn't totally scale- Like, there was- Well, you're serving, was it... You're serving too bigger customers to start with? Yeah. Is that the problem? Yeah. And, and- And all with different needs... 6:59 and they were too disparate. Yeah. You know? At the enterprise level, the needs are so disparate, and they're so... 7:05 And also, if you're i- i- in that realm, if you're trying to replace an agency, they're used to asking the agency for anything they want- Yeah... and getting a really custom- Your bespoke. Yeah... 7:14 kind of bespoke service. And when you're trying to replace that with automation, they still expect the level of bespokeness. Yeah. And so we actually, 7:22 we went away on a founder summit for a few days, and on that trip we, we... There, there was a talk there. It was a Squarepeg summit, where there was a talk from one of the founders of Duolingo. Yeah. 7:33 And he was great, Louis. And he said, um, "Hey, you've gotta always take the long view." He was talking about building Duolingo, which is one of the greatest, like- Yeah... products- Yeah... built in software. 7:45 He was talking about just the, the craft and the time that went into building that product, and not at, in the early days worrying about go to market, and he kept... And he would said, "Always take the long view." 7:55 That's- Yeah. And I was like, as soon as he said that, I looked at our co-founders and I was like, "We're not taking the long view." [laughs] Yeah. You know? 8:01 [laughs] Like, we know, we know what we wanna build here, which is we wanna be able to automate advertising end to end for customers- Yeah... and, you know, where humans don't need to be in the loop anymore- Yeah... 8:10 and use software production to drive- To do the work. Yeah... the AI algorithms- Yeah... in Meta. Um, and so we said, "This isn't working." 8:18 And so we came back and- But you've gotta have, to get to the long, turn the long view's kinda right, but you've gotta fund... You gotta be able to get- Yeah... to the long view. So sometimes... 8:27 So how do you balance, you've got a product in market, people are buying it. Yeah, yeah. And we- You basically fired your customers, from what I understand. We did. 8:33 We had just close to a million dollars in ARR at that point. Yeah. And so it was, you know, it was looking good. Yeah. We were probably six months in. Um, but I ju- could just see the unit economics weren't gonna scale. 8:46 It was kind of like, it was a bit of vanity, a bit of vanity metric there. Like the- Yeah... the numbers looked good, but I knew underlying, the product required a lot of people running around doing things- Yeah... 8:56 manually. Um, yeah, so we just knew we had to, we had to step back- Did you have alignment?... get off the hamster wheel. 9:06 Did you have alignment on that across the founders and investors and- Well, look, Squarepeg were amazing. Um, we went to them and said, "Look, we actually wanna fire all the customers." Mm. 9:16 "And we wanna, we wanna take the technology we've built, we wanna start the product again. 9:21 We've got a lot of good tech, but we wanna re-engineer the user experience end to end so our customer can sign up, no humans in the loop. Our AI can learn everything about their brand. 9:31 And after onboarding, they could be publishing their ads live." So really fast time to value. Really fast time to value. Yeah. 9:37 Like, within 15 minutes, you know, there, there's no reason within 15 minutes you couldn't be up and running- Mm... and no humans are required in the loop. Yeah. 9:45 And so that was our, that was our goal, and that's what we knew we needed. And so you've... Yeah, okay, you stopped them. Well, you're using some of what you've built. Mm. How long to get V2 ready and into market? 9:58 So it was- And you had to raise more. Did you raise... You raised- Squarepeg were great. They said, "We back you. We think that's a great idea, and would you like some more money?" Right. Which was amazing. Yeah. 10:06 Like, I mean, you can't ask for anything better- Yeah, that's perfect... from, from your seed investors. Um, so then we set a goal. We said, "Look, we're gonna give ourselves six months." Mm. 10:15 So it was liter- I think it was January 1. So six months of essentially no revenue growth. Yeah. Probably revenue- No revenue... declining. We, we'd canceled all our contracts. We refunded a bunch of customers. 10:24 So essentially not o- not only, you didn't just cede it a million, you went- We gave money back. Right. Yeah, yeah. Um, but it freed us. Yeah. It freed us from 10:34 all of the, you know, demands, the little insidious things of having to continually- Yeah... make changes for customers instead of focusing all our attention on, on building. Um, 10:45 yeah, so we, we basically said Ju- it's, it's January 1st, we're gonna have six months. July 1, we're gonna relaunch. Mm. 10:50 And I, the thing I've learned from building software companies is, um, there's nothing like a deadline to make things happen. Yeah. Um- Yeah... everyone in software, you know, my background software- Likes to be agile. 11:01 Yeah, yeah... they all say we can't set dates. Yeah, yeah. Dates are terrible. It's all, yeah, agile, scope is adjustable. But, like, in the end, if you set a date, things happen 'cause- You gotta move towards it... 11:11 you find smart solutions. You, if you have to work to a date, you will find a way, and doesn't mean, that just means everyone works harder. Sometimes it does, but it'll often mean- Just-... do we really need to do that? 11:21 ... people, you have the right priorities, yeah. Is there a better way of doing that? Yeah. Yeah. So, um- So six months until you hit that- Six months... hit that deadline. 11:27 And, and so we incubated over those six months with... So first what we did is we, we went to a bunch of customers. We did, like, 50 customer interviews. Yeah. 'Cause we knew there was a problem. Mm. Like, 11:39 customers can't make enough ads to feed the algorithm. Yeah. So anyone who advertises has this problem that, like, you know- Yeah, they wouldn't be able to test it, yeah. Yeah. 11:45 So it's, like, it's, it's a universal problem. Yeah. But then the question was, like, who- So these are ads that, being able to get, and just for those who- Yeah... 11:52 haven't advertised, so you, you, the ability to produce lots of ads- Yeah... will enable you to test those ads in, through Meta or whatever platform it is. That's right. 11:59 So Meta have automated away all of the bidding and targeting now. So you can't, you don't go into Meta and say, "I wanna target You know, men between 25 and 30 in this area Yeah You don't do any of that anymore. Yeah. 12:13 That's all managed by AI. Okay. The only input you have as a customer is the actual ad content itself- Right... the creative, and then Meta's AI will find the audience that resonate with that ad. Right. 12:24 So it's like what you need to do is feed it a whole lot of concepts. Yeah Like, different ways of bringing your product to life. 12:31 Like, if you're an insurance company, um, selling maybe car insurance, you, like, you, it, you might wanna, uh, you might wanna explain the value of that car insurance through the lens of someone who's camping- Yeah... 12:46 for example. Yeah. And anyone who's a big camper would really resonate with that. Yes. Like, "Oh, it's annoying when you're out and this happens in your car and you're in the middle of camping," and you know. The, the... 12:55 Yeah. So not everyone [laughs] who buys car insurance has the same needs or the same context. Yeah. So you wanna, like, tell your story in many, many different ways. Yeah. 13:03 So for campers, for this person, for that person. Yeah, for people on their commute- Yeah... to, to work every day, for, for people who drive long distances. Which is really hard to scale up if you don't have the tools. 13:12 Yeah. And there's so many. Um, um, you know, or even like, you know, think of a company like an Uber or a Didi. Like, um, 13:21 you know, there's many reasons you'd take a, a ride-sharing service, but you wanna talk to people who are into different things. Yeah. Like maybe, is this something you do to get to the footy? Yeah. 13:28 To get to the footy on time, or you do it- Yeah. To pick up your kids or... when the train fails and- Yeah... you need to get into work As a backup. Yeah Like, there's different ways of telling the same story. Yeah. Um, 13:37 and so if you tell the story in the right way, Meta will find all the people that story resonates with, put your ad in front of them, it'll convert better. Yeah. 13:44 So what you need to do is have a, a very large volume of ads that it can find the right people. Yeah. So that's kinda how you do creative variation, sorry, like targeting and, and bidding now at scale. Yeah. 13:54 Um, so they call that creative variation. Yeah. So we knew that was the problem. 13:57 The, the, then the question is if the TAM or, for your company is basically any company who advertises, which is most companies on Earth, like who are you actually selling it to? Yeah. So our, our big question was like- 14:09 So who's, what's your ICP? What's our ICP? You're really trying to work out who your, who your customer is. Yeah. So we started by going, "All right, let's go and talk to 50 target customers." Yeah. 14:18 We figured we wanted to probably do something in the, like, e-commerce space 'cause they're big, um, advertisers- High volume... on Meta. Yeah. Yep. So we went and interviewed 50 customers- Yeah... 14:28 manually, recorded it all- Yeah... synthesized it, and we noticed this one, um, theme over and over and over, which was, "I don't like AI. I'm scared of AI. Um, I don't want to put AI slop out there- Yeah... 14:41 from my company." Like, you know, you think you're a marketer, you're building ads for your brand out there in the world, the last thing you want is it to look AI generated. Yeah. Um, and, uh, I, and, and I've got 14:56 folders and, like, Dropbox drives full of video shoot assets and photo shoot assets- Right... that I can't afford to repurpose. Yeah. 15:04 So the challenge was actually, hey, uh, it's not- So, so using, I've got all the, I've got all this content that I wanna be able to use- All this content... 15:10 but I don't want it to look AI generated, so it's gotta be high quality. Yeah. So, like, I don't need to AI generate. I've actually done a, I've paid for so many photo shoots- You need to pull it together then... 15:17 and video shoots, but it's all sitting there on drives, and it's too expensive for me to pay video editors to sit there and manually- Yeah... recut that into ads over and over and over. Yeah. 15:25 So we noticed that same feedback over and over, so then we said, "All right, well, let's build the first version of this, this new Cutable- Yeah... which we call Cutable 2.0." Yeah. 15:34 Um, let's ingest existing assets- Yeah... and recut that into new stories, new ways of, like, connecting with the audience. Yeah. 15:42 So let's connect what they have into something, into videos that can run on Meta that tell a variety of stories and we can publish a volume. Yeah. So that's where we started. Yeah. Um, so yeah. In six months? 15:53 You gave yourself six months? Did you have that- Gave ourse- six months, and we- So you'd done all the interviews, so then you start, you know what you're gonna build or you have some view. 15:59 And we, and we took on, like, 30 alf- we called them alpha customers. Yeah. So customers that- [laughs]... weren't paying us- Yeah... who were, like, just helpful. Yeah. Like, and we were kind of 16:10 building for them during that time, getting it in front of them. Yeah. I, I'm- And, and you're ho- at that point you're hoping that they ha- will have common needs. Yes. 'Cause otherwise you're back where you started. 16:19 Otherwise you're back where you started. Um, and I'm not a big fan of doing this, so this is why I wanted to have this six-month period and, and, like, doesn't matter, come hell or high water, we are launching on July 1. 16:31 Yeah. Because I'm not a big fan of, like, relying too much on customers who aren't paying you. Yeah. Because I think there's a lot of people who are very well-meaning- Yeah... 16:39 and wanna help out, but ultimately you don't actually know if they're not- You don't know if they're gonna pay, you don't know how much they're worth. Yeah... 16:43 you don't know if they're willing to pay to solve this problem. Yeah. They might say it, but if you actually then go to charge them and they have to pay on a monthly basis- Yeah... 16:51 is this problem really enough of a problem for them to pay for? Yeah. And so you wanna make sure that it's people in that category that are giving you the feedback- Yeah... that you're using to craft your product. 17:01 Um, so that f- first six months was hard. We definitely had some great customers who leaned in, and some were absolutely ideal ICP, like when we look back now, and others weren't. Um- Yeah. 17:10 So those, so those 30 customers weren't ne- they didn't necessarily all become your customers- No. No... from that point on. 17:14 And you don't wanna build y- the danger is, where founders do this early, they get stuck building for s- a company that ultimately would never actually buy their product. Yeah. 17:23 Because they're getting it for free, and they, they're me- they're well-meaning, and they think, "Oh, it sounds pretty good. Yeah, that sounds like it could be useful." Yeah. 17:29 But, like, if you actually make them pay for it, will they pay for it? Well, and also sometimes the person who has the time might not be the person that- Mm... is gonna be your best customer either, so. Yeah. 17:36 Yeah, exactly right. Yeah. Um, so we launched on July, maybe it was July 10th. Yeah. I think we were 10 days late- Yeah... but we did pretty well. [laughs] Yeah. And I think we made one sale in that month. Right. 17:49 [laughs] Um- So at that point, what are you thinking? Because, uh, it- Are you thinking pivot's worked, or you just... No, I think we thought we were onto something, but, like- Still slow?... 17:56 the product didn't do a lot. Right. It made one kind of ad. Um, and so you could at that, we started with brand ads because we knew it would be easier to execute. Mm. But, like, you couldn't make a product ad. 18:08 So if you had a- Mm... if you were trying to sell a product you couldn't make that with Cutable, which is, like- We're talking about brand ads... pretty insane. 18:12 Like- So you do a brand, but you couldn't do a- Yeah, "Hey, come buy this product we sell." Yeah. 18:16 So you could talk about your brand and we could say, like, we could do, like, stories about your brand, but you couldn't actually sell a product- Yeah... 18:22 which, which is basically all that D2C brands are trying to do [laughs] Yeah, yeah. So that was, and maybe, we probably made the wrong call there. Yeah. 18:28 Um, but we thought, hey, brand ads will be easier to execute on first. Yeah. 18:33 Um, and then really over the next three months, July to September, we just spent three months getting punched in the face, 'cause we were just getting on demo calls, trying to sell this thing- Yeah... 18:43 onboarding new customers, having some of them love it, having some of them hate us, you know, hate not us, the product. [laughs] Yeah. You know, 'cause it was so early. Yeah. 18:52 Um, but it was just such an important period of time of, like, refining. We've got this core technology, now we're, like, what are all of the things that are wrong? Yes. 19:01 And I would say during that period we, we learned all of the things that were needed to make the product better, um, to just get it up to a- Yeah... 19:10 a level of, like, delivery that, hey, this solves the problem for customers. Yeah. Um, so yeah. And you're on the tools personally. I was, yep, yep, yep. You were. At that period you were. Not... 19:23 Yeah, I was a bit, not- But you got a dev team- Not every day... of, like, five people or thereabouts at the time? Yeah, yeah. Yeah. And they're just- I think we had three... three at the time. Three or four, yeah. 19:31 Okay, so you're moving fast with- Yeah... obviously AI helps a lot in that. Yes, yeah. We, we intentionally kept the team very small. Yeah. Um, I love our head of engineering. 19:41 Every time I would go to him and say, "Do you wanna hire another engineer?" He'd say, "No, it'll slow us down." Slow us down, yeah. We, like, "If we wanna move faster, we need to do more with less." 19:48 Like, that's his attitude, and I think it's- Yeah... absolutely proven to be the right call. Yeah. Um, uh, because it's very easy, it's a very easy trap to fall into, to say, more hands on keyboards means more delivery. 20:03 Yeah. But is it, are you delivering the right thing, or are you just having more conversations- Yeah... about the wrong thing? 20:08 As a little side note, I was speaking to one of our engineers at Firmble, and one of our senior engineers who's been with us for a few years now, and we're talking about the speed at which we're building. Hmm. 20:17 And I meant the company more generally. I, uh, and I kind of was saying to him, "I, I think we're at this point where I think it'd be very hard for another player to catch up to what we're doing." Yeah. 20:26 'Cause we're, we're, we're getting this iterative- Yep... you know, recursive nature of how we're building things. He understood it as being people, and he said, "Yeah, it's very hard to add somebody into the team." 20:37 'Cause he's, he's- Mm-hmm... your classic 10X engineer, where he's- Yeah... running agents, doing all his- Yeah, yeah... you know, and he's got a team of half a dozen people. 20:43 But for him to a- for him to add a new person- Yeah... they need to be able to keep up with what he's already doing. Yeah. Otherwise, he takes 10 or 20% of his time where he's not running his coding agents- Yeah... 20:53 which slows him right down. Yeah, yeah, yeah, exactly right. [laughs] So he sort of relating to- It's so true... and, um, once we- Yeah... 20:57 realized we were talking about different things, but I said, "Yeah, tell me more about that." But just, it, it shows you, you just can't plug more people necessarily into a team. 21:03 You get to a size that works, and to step up to the next level is- I'm also finding with AI, you need engineers with imagination- Mm... and commercial pragmatism. It's like that blend. You need someone who can go, 21:15 given what it is today and how customers are reacting, I'm going to imagine the next evolution of this thing. Yeah. It's not what the customers are asking for, but I think it's what they need. Yeah. Um, 21:27 and you can't- I think product and engineering has become- It's very hard to just hire someone in to do that. Yeah. Like, they don't- I think actually product engineers become one- Yes... 21:33 I mean, it should always have worked as one team. Yeah. But invariably you had product people and engineers. Yeah. Now that's blended. It's very much blended. 21:40 We're having these constant conversations around what is the role of a PM versus an engineer in- Yeah... in this world? It's very, it's very unclear now. Yeah. Yeah. Um, yeah, so then we, so we get to September, 21:54 and then I think what was really foundational for the company- So we're talking September '20- Last year, 2025... five? Yeah. Yeah, so about a year ago. Yeah. 22:02 Um, we're now starting to sell this thing at a reasonable pace for the- Yeah... first time. Um, and now it became a question of, well, who is the right audience? Yeah. Like, what is the ICP for this? 22:14 Because everyone on earth wants their ads to perform better. We've now got a product that can deliver ads at scale. Yeah. But who is the right audience for this? 22:24 And so what we spent the next six months doing was just selling it honestly to, to anyone who wanted to buy it. Yeah. Which at the time felt frustrating, 'cause everyone was like, "Well, who is the ICP?" 22:35 But we didn't have enough data- Yeah... to know who- So you hadn't l- you hadn't tried to land it at that point? You were just saying, "Let's just see what happens." It was, it was just... 22:42 We were having constant conversations. Yeah. But when I look back now, I re- like, I'm not sure it was totally in- intentionally stated at the time, but we kept it quite wide. Yeah. 'Cause we still didn't know, is it, 22:55 is it, um, like even, like what i- is it... How do you even cut it? Yeah. Is it, like, the team size? Yeah. How do you work out what, how do we even cut your ICP? Is it how much they spend on advertising? 23:06 Is it their team size? Is it the right, what channels of advertising they do? Yeah. Like, what, how they advertise? Is it industry? So we, we, we kept kind of selling it to anyone for six months 23:19 so that we could bring on a wide variety of customers, and then look at the data and cut it different ways. So who's using it the most, who's- Who's using it the most, who's, who's better converting. Yeah. 23:30 Who's staying, not churning, you know, of like- Yep... obvious things like- When we do a demo, what percentage convert to paid- Yeah... or to doing a tr- we did trials for a while. Yeah. 23:37 Um, when they come on, who passes month one, month two, month three? Yeah. What does churn look like? Um, and so that six months was, like, foundational. We brought on a lot of customers. 23:49 It was frustrating, 'cause it felt like we were losing a bunch of customers as well- Mm... but it was because we were working out who- Yeah... who best to sell this to. And it was probably early in 24:01 2026 where we started cutting the data and realizing the best way to cohort our customers and find, to, to understand how ICP correlates with all the things we care about, was how much they spend on advertising every month. 24:16 Am I s-- A- And mostly Meta is the channel for- Mostly Meta. Yeah. Yeah, we do TikTok and starting to do Google now, but- Yeah... um, you know, most of the D2C, we sell to D2C, like- And they're mostly on-... 24:26 direct-to-consumer brands- Yeah... and they're mostly on Meta. Um, and so we found, hey, um, customers that spend $10,000 a month or more on Meta 24:38 are, you know, three, four times more likely to convert from a demo to paid. Yeah. Like, I, I can't remember the exact numbers now. Maybe three times- Yeah... less likely to churn. 24:49 Um, and so we started segmenting, um, those markets. Is that one just b- back to biggest customers again, or is it certain types of- No, 'cause it would actually then tail off. As they got bigger, it would drop. Right. 25:00 S- so customers who spend, like, 10 grand- But they're small... to- So as, as a proportion of their advertising, it's the biggest, it's a big chunk. Is that the... Yeah. So it needs to be... 25:11 So what, what we're finding is we had a lot of, like, founder-led brands- Right... or customers. It might even be big companies, but they've never done meta advertising before. Yeah. 25:19 So often they'd be, you know, $100 million revenue companies- Yeah... haven't done advertising before on Meta. Yeah. Come to us, "We wanna get into meta advertising. This looks like a great solution." Very rarely worked. 25:30 Yeah. Because if you haven't done advertising, meta advertising before, you don't know the foundations of it. And often, when your ads don't work, they would say, "Hey, the ads don't work." Blame you guys. 25:39 "It must be your s- solution." Yeah. But actually, the landing pages aren't optimized, checkout's not optimized. Yeah. Um, maybe not putting the right budgets behind it. Yeah, so essentially- Yeah... 25:49 you're getting companies who already have got meta to work for them. Mm-hmm. And you're improving. And volume is the problem. And volume is the problem, yeah. Yeah. 25:55 Or, or with, like, a lot of the founder-led early stage brands, they don't even have product market fit. Yeah. Right? And so they're trying to, like, advertise on Meta while they're working out the PMF. And- Yeah... 26:06 so yeah. We, we learnt over time, um, you know, brands that spend between $10,000 a month and a million dollars, $2 million a month work really well on CuttLE. 26:18 When they go beyond that, often you'll find the organizations are much more complex. You know, you've got- Yeah... back to, like, you've got 50 people- Yeah... 26:26 involved [chuckles] in the process, and that it's not clear where CuttLE fits in that process. Yeah. Yeah. So what percentage are you then, 26:34 your contracts end up representing what percentage of their advertising on Meta, roughly? So that if, if they're spending- If, yeah... 100 bucks, are they spending 10 bucks on you or what? It will, it does vary. Yeah. 26:43 Um, the typical rule of thumb in advertising is if you spend, it's 20% should go- 20%... towards creative production in total. 26:52 But you'll be less than the t- So if you spend a million dollars on ads- Yeah, you spend 20... 200 grand on the production of those ads. Yeah. But visually, you reduce that. But we're obviously not all of that. Yeah. 27:01 We, we want to be all of that. Yeah. But we will get there. [laughs] Uh, so it'll be anywhere from kind of 1 to 15%, we'll say. Yeah. 27:09 And it depends how much, um, they're making their, all their meta ads with us versus a portion of their meta ads. Yeah. And that comes down to a little bit what kind of ads they make. Yeah. Yeah. Um- Yeah... 27:20 but yeah, we generally sit in that sweet spot. Yeah. And that's how we end up thinking about pricing as well. Yeah. What... So when we're able to segment the customers by 27:30 meta ad spend, we're also able to say, "Okay, well, um, you know, what are, what would we represent as a percentage of their spend?" Yeah. "What, what is a reasonable amount that is value to money for that customer?" 27:45 Um, and so you can kind of proxy it that way. Yeah. Yeah. Yeah. Which allowed us to raise our ACVs, so we 8X'd our ACVs this year from, so our average sale price at the start of the year was, 27:57 yeah, eight times less than it is now. Yeah. Yeah. Yeah. That's amazing. Yeah. Okay, so you've, um, you've retooled the product. You've found your ICP. Yeah. 28:06 Before we move on to next stage, which you, you just raised capital, um, and we'll talk about that. But- Mm... 28:11 when you thought about building the product, to talk a bit about how you think about what you, do you train your own models? Do you use, is everything from, you know, front team labs? 28:21 How, how, how are you sort of thinking of- Yeah... not on the coding side, but more on the, the models your, your customers use? We, we use a ton of different models. Mm. 28:30 Like, our, our software will use the best model for the best job. Mm. And we're constantly changing them. Um, we don't train our own models, but we add, um, a whole lot of context on top of it. Yeah. Yeah. 28:43 Engineers can swap the models in and out, presumably. Um, but then- Yeah... you're also putting essentially a, a layer on top of that, um, and- Yeah... sort of like AI 101 in a sense. 28:52 Um, so walk us through, so you've, you know, you pull in the model. Yeah. You have context that you have as CuttLE and also customer context- Yeah... uh, that you bring together to enable- Yeah... 29:04 the software to produce an ad. So- Yeah... um, so all that compute, so you're, you're running all of that. You're managing when it uses the, you know- Yeah... 29:14 different models and when it then brings in the context that it needs. Yeah. And it's a smart workflow too, right? So it's not like one thing, but we will, we will ingest all of our customers' assets, for example. Yeah. 29:26 So we will watch every millisecond of every video they've ever shot with AI. Yeah. And we'll know exactly at that point in time in this video- Something happened... there's this shot. Yeah. The customer says this. 29:38 Um, we know general themes. 29:41 We'll a- be able to, like, rerun classifiers so we can start to aggregate, hey, there's generally in their, in their photo shoots, they'll generally do industrial warehouse with this kind of lighting. Yeah. 29:52 They'll do family in the park, vivid life, you know, like life scene. Yeah. And so you start to learn- So you're describing, you're describing their brand and, and how, and what the ads look like. Yeah. 30:02 You're building up that- We're building up a lot of context for a brand. Contextual, yeah. Yeah. To- tone of voice they use, and then how they might apply that voice differently to different audiences. Yeah. 30:11 'Cause we learn all their different audiences. Yeah. Um, and then, you know, there's just the aesthetic stuff. They're like different color themes. 30:18 There might be their base color theme, but then when they do a sale, they always do it in this, you know- Offset kind of like- Right... you know, everything switches and the colors are inverse and they use a blue. 30:28 And so we learn, like, all the different color themes as well. Yeah. And so there's a lot of nuance to a brand. And so, so a customer signs up, how long does it take? 30:37 Presumably that can happen reasonably quickly within hours- Yeah... days, that you can ingest all that content. So, so customers- Yeah... I sign up, I start, I, I wanna use a product. Yeah. 30:46 I then want, connect it to my, all, all my files, all my videos. Yeah. We'll, so we'll first do a scan of all their public stuff, website- Yeah... 30:53 Meta Ad Library- But can you do that almost real time when they sign up and- Yeah, we do that before we- Before your sales call... do a sales call with them. Yeah, before your sales call. 31:00 So we've actually built a whole Cutable account for most of our customers before we do a sales call. Yeah. Um, and, 31:07 you know, depending on the customer, we might actually show them that account or not, depending on what we think, where we think they're at. Yeah. 31:12 Um, yeah, so we build that first, then when they sign up, we ask them for more context. Yeah. So, you know. Give us your videos, give us this- Give us your videos- Yeah... give us your assets. Brand guidelines. 31:22 Brand guidelines. Yeah. Um, connect your Shopify and we'll s- Yeah... we'll bring in all their products, all their product imagery from Shopify. We'll learn everything about them. They can now connect, 31:32 um, Klaviyo, Google Analytics, so we can see what customers are searching for on Google when they hit that website. Yeah. What a customer's interested in. Um, so we have all different MCP connectors now as well. Yeah. 31:44 Um, so we build this whole profile of the brand. And all this- And have they ever had that before? Uh, uh, I presume a lot of these customers have never really had built that context themselves before. 31:53 No, no, they haven't. And they're often... You often hear from customers like, "Wow," like, "Yeah, I never thought about my brand that way, but that's right." Yeah. "That is how we talk." Yeah. 32:01 Or, "That is what we, that is how we show up." Yeah. [laughs] Um, which is really nice. Yeah. I really wanna build in, like, a reverse pitch. 32:07 When you go to an advertising agency, the first thing they do is a reverse pitch- Reverse pitch, yeah... where they tell you- Yeah... what they've learned about you, and it's a great way of building trust. Yeah. 32:14 I wanna automate that- Yeah... and do an AI reverse pitch, 'cause it's basically what we do anyway. Um- Yes... but yeah. Um, yeah, so customers come in, and then, um, 32:24 yeah, we, we, we connect all that additional infrastructure, we learn, and then we start making ads for them. And, and what we've rolled out now is, um, a whole new system called Ads That Learn. Yeah. 32:33 And that's where we start. We're now being proactive and rather than customers... Customers can come to us, and they always have, to say, "I want an ad." Let's say, "I want a f- Father's Day campaign- Yeah... 32:43 for this product highlighting this," we'll go and do that. 32:46 But now we're actually proactively going to customers each week and saying, "Hey, based on what we know about you, we think these are the ads you should run this week." Yeah. 32:53 "And we've actually built them for you already." So you're continually running, are you continually then almost running in the background? Yep. Yep. And so compute cost, then, how do you manage? 33:02 'Cause that's a lot of inference at that point. Yeah. Are, are you using- It gets expensive. [laughs] It gets expensive. Is it all... Are you using open weight models? Um- Uh-... 33:10 a blend of different open source and- We use a blend. We mostly use the frontier models to get- Just 'cause they're just better? We'll use the cheaper ones for the bulk tasks. Yeah. 33:18 Uh, like what I say to the engineering team is like, "Don't worry about cost when you're inventing." Yeah. "Just get it, get it right- Yeah... 33:26 and get the best-in-class solution to a customer and validate with them whether they love it." Yeah. "And then if they love it, we can worry about cost optimization." Yeah. Yeah. 33:34 And are you optimizing on behalf of your, all of your customers, or are you allowing them to choose? No, they don't choose the models. They don't choose, no. No, no. It's all invisible to them. Cool. 33:41 But we do that behind the scenes. Yeah. 33:43 But if you, if you listen to the way I describe it, there's so many different processes happening there across the interactions that customers have, like sucking in their assets, watching their- Yeah, so some of that you could use on, um- Using all the connectors. 33:54 It's, you'd use a mo- a different model for all of it, right? Yeah. Use a cheaper model for ingestion maybe- Yeah... 33:58 and then in generation it might be, you know- Ingestion and classification you can use cheaper models these days- Yeah... because the, the... 34:04 Like two years ago you needed the frontier models, but now you can get- Yeah... a really good outcome with the cheaper models. Yeah. Yeah. Um, 34:12 uh, but when we're talking about ideation- But with purely for generation, ideation, you're gonna use a frontier- Ideation, we wanna use frontier models- Yeah... because you wanna be as creative as possible. Yeah. Yeah. 34:20 Um, but then, you know, sucking in things from other systems. So an example would be we talk to Shopify constantly. Mm. And the moment, you know, let's say you, you're starting to run out of stock of something. Mm. 34:32 We'll go, "Oh, you're getting low on stock on this. Do you wanna just sell through all of the rest of it?" Yeah. "We've built a hurry limited stock campaign for you. Here it is, it's ready to go. 34:42 You can just push that out and just sell through the rest of it." Yeah. Um, so that kind of, like, rote learning we can bring in, we can do that with a cheaper model. Yeah. 34:50 But then the actual ideation of, well, what, what would the ad look like to bring that to life and- Yeah... and get people to click on the ad, then the actual idea for the ad, you'd use a frontier model. Yeah. 34:59 But then maybe the construction of the ad you'd use a- Yeah, yeah... like a less, um, less expensive model. Yeah, so how many points are you ma- Yeah... 35:07 how many processes, just roughly, are you making decisions as to which model to apply? It sounds like dozens of different points. Oh, yeah, dozens and dozens. Yeah. Yeah. Yeah, yeah. 35:14 'Cause these are like long-running workflows and- Yeah... and it's agent tech and so- So you've got your own... 35:19 You essentially kick off, yeah, you're kicking off all those build processes and the agents are running the agents. Um- Yeah, yeah... 35:24 and they're working out what to use at different- Yeah, and the different agents will even be different models behind the scenes potentially. Yeah. Yeah. That's as far as my knowledge goes though. 35:32 [laughs] Yeah, you're really pushing me to the edge here. Carl's probably watching this going, "That's not true." Well, that's not true. Yeah. [laughs] Close. Close, but n- not quite. 35:40 Um, okay, so you've, you, you're, yeah, so you, you're out now, you're getting great traction. Uh- Yeah... so talk to us about the, the current, or the series A you just did, how you thought about the round, um, why now? 35:50 Presumably you need the money- I, I-... but it's more than just that of course... 35:53 I would say, like, what happened from kind of March this year to September is we, we had enough data to really refine the ICP and focus the business. Yeah. 36:05 And so every few weeks we'd made further refinement decisions- Yeah... and really narrowed in. And every time we did that, all of the metrics of the business got better. 36:12 Yeah, so metric getting better, you're getting better- What, faster conversion? Faster conversion, yep Higher deal size. Yep. Um, better retention And inbound lead generation. 36:21 We could focus more and more on the right kinds of customers- Yep... so we're getting better conversions, and then scaling up more of the right customers. 36:26 And your main ex- your main, uh, channel from a demand point of view, I presume, is Meta? Is Meta, yeah. Okay [laughs]. We're, we're lucky in that perspective- Yeah... 36:34 that all of our custom- like, generally in the past people- So your customers are sitting on Meta, you're using Meta to get to them, to show them how they can- Yeah. Yeah. Yeah. 36:42 Um, it's not the only way, but it is, like, it's, it's a, it's a great way to reach our customers. Yeah. Yeah. Yeah. Um, yeah, so business has been getting, yeah, really gone from strength to strength this year. 36:53 Yeah Really started flying. Um- An acceleration in, in... And you've built a go-to-market team? Yep. So we, we found it was best to sell this product, um, with real humans. Yep You know, strong human- So you think-... 37:05 led sale, not, not product led... the most AI native businesses still need people. To talk, uh- Yeah, yeah... something comes up in our business, obviously, when people talk about, do you need a salesperson? Yep. 37:15 Our view is, well, if you've still got a human at some point involved- Yeah... you probably still need a human in the sales process. But you of course automate and make them way more productive. But- I think so. 37:25 And I think what, what we found, so we did, we did an experiment for a month. We sold self-service, so customers could just- Mm... come in, sign up online, no salesperson in the process. 37:36 This was earlier when the product was less good, but even still, I think what we found is with AI, customers want a very quick result, and they're very quick to judge. Yeah. 37:48 So if you're, if you're fully AI automating, um, their, to the point of their first experience- Mm... um, if you get anything wrong, they'll- They're off... write you off- Right... completely. Yeah. 38:01 They'll just be like, "This is AI slop." And so what we were finding is, hey, Cuttable sometimes would get it right first time and sometimes wouldn't. Mm. 38:11 But when it wouldn't, it just requires a, a little bit of feedback and a few tweaks, and we're in a, we're in a great place. And you can control that better if you've got a person involved. Yeah. 38:19 And so you need to be able to, like, have that discussion, like, how would your brand typically show up this way? Mm-hmm. And, um, and when you have that discussion, people are very open. 38:29 You go around, you might spend 10 minutes on it, you kind of tune the system together, and then it's, it's good to go. Yeah. 38:35 Um, but when they were doing it themselves, um, it was almost impossible to rescue the customer after they had a, a poor first experience on their own. Yeah. You'd reach out and say, "Let's get on a call." Yeah. 38:45 They're done. You're like, in their mind, you're AI slop, which is what our customers are against. Very cool, yeah. Very fun, yeah And that's what we're against. Yeah. 38:51 But we, it was very hard to convince them that actually this thing is real- can really nail your brand. Yeah We just need a bit of feedback. Yeah So you need a sales team, so you're building that out. We need it, yep. 39:01 Yep. Yeah, so we, um, we built that out here. We, we opened in New York earlier this year, so we've got a sales pod- Yeah... out in New York. Um, covering the East Coast of the US, we've just opened in San Francisco. 39:13 Yep. Um, so for us, scaling the name of the game is hiring more AEs and- Yep... then hiring, um, customer success- Yep... and implementation folks to help onboard customers. Yeah. Um, yeah. Mm. 39:27 Um, yeah, so I think off the back of, we, we saw really strong growth, um, from about March through to now. Yeah. Um, through to, what is it, end of October right now. Yeah. 39:37 Um, and off the back of that, yeah, raised a really strong Series A with AirTree, which we're- Yeah... very excited about. Um, yeah, just closed that out and announced it recently. Yeah. 39:50 And AirTree knows you from the Cloud, Cloud Guru? Yeah, yeah, yeah. I've- They were investors back then. They were investors, yeah. James, I've known, I actually have known James since he was at Accel in London. Mm. 40:01 That's where I first met him, um, when we were raising Cloud Guru's Series A. Yeah. Um, I met him there and he said, "Hey, I'm actually going back to Australia to join this fund called AirTree." And so- Mm... 40:11 um, that's how AirTree got involved in, um- Yeah... in Cloud Guru. And, um, they were part of our B and C rounds at Cloud Guru. And yeah, so I've always, I've always had a lot of respect for James. 40:21 I think he's one of the best investors in the country. Yeah. Um, so yeah, very excited to, to have them lead our A. Yeah. Yeah. Just one other question. 40:29 I'm, I'm just jumping back a bit, but on the sales side, it's always, we're hiring, as you know, Firmwall's building, we're building in the US as well. Yeah. 40:35 How do you find hiring in the US, hiring salespeople and building the team there? I think it's... I don't know what you find. I'm interested. It's hard [laughs]. I find it hard. Yeah. 40:44 I think we're, here people know who we are. Mm. Like, we've got a bit more cachet- Yeah... whereas over there, which it's just like, "Who are you?" Yeah. Like, "Haven't heard of you." 40:52 Um, and Americans are very good at talking, talking a good game. Yeah, salespeople all sound great [laughs]. They, they all sound amazing. Oh, they interview well. Yeah, they do. Different Americans, yeah. 40:59 They interview so well. They've got so much confidence. Yeah. Um, so yeah, it's really tricky. Yeah. It's really tricky. 41:05 Um, but we've, we, yeah, we've had, we've had some good success so far, but we've definitely had misfires as well- Yeah... where we've hired the wrong people. Yeah. Yeah. What about you? Yeah. 41:15 Uh, Firmwa- well, Firmwall is, the one benefit we have there is that salespeople kind of get our product. Yeah, yeah. Um, so- Yeah... that makes it a little easier. I mean, at A can even go way back. 41:23 We, our first 10 salespeople in the US, only one made it. Um, so- Oh, really?... it was really hard. Yeah. But that was sort of more enterprise sales, so we're, they were expensive people. 41:32 We needed them to understand construction, um- Mm... and understand how to sell a software product. We're getting that intersection of, you know, SaaS and construction was hard. 41:40 Uh, not that they needed to know construction, they just had to know enough to be able to have a critical conversation. Yeah But it was, uh, it was very hard. 41:46 Uh, Firmwall's been easier because we're hiring more junior people. Okay, yeah. Um, they're not enterprise sales, more so your SDRs and AE types. I mean, more, more so mid-market AEs. Yep. 41:56 Uh, and having, selling into go-to-market is, you know, we're, people get our product. Mm. So it helps that a salesperson broadly understands how they might use our product. Um, but it, it's hard. It's hard. 42:07 And, yeah- I would say that- And we've got a team in San Francisco, and there's just so much happening there in San Francisco at the moment that, yeah, getting- Mm... 42:13 getting attention is, is, um, difficult, but- I, I think- Yeah... I think one of the things that worked for Cloud Guru is our first- Go to market hire was in the US, based there and had a network. Yeah. 42:24 So that helped us bring in the first five or 10 through- Yeah... his network, which really got us moving. Got it going, yeah, yeah. 42:31 Whereas what we've done with Cuttlebot, we've got our head of revenue here, and he's- Yeah... going back and forth. And, um, it's just... And there's been a lot of advantages of that too, 'cause we've- Yeah... 42:41 built it all here to start with, but it means we don't, we haven't had that ready network, so it's just maybe been a little bit slower. Um- Yeah... 42:48 but I think it's now really starting to, like, now we've got good people there, the networks are starting. Yeah. Yeah. Yeah. 42:54 Yeah, and the other one will be, we're building our marketing team, not just our sales team, and that's, yeah, again, you know, just sticking out as a brand in the US, particularly in San Francisco at the moment, is just hard. 43:03 Um- Yeah, yeah... but I think, um, you know, leaning into what we have that's different, um, our culture and, and even the fact that, you know, not, not hiding the fact that we're Australian. 43:13 Like, there's, there's a lot of people that- Yeah... there's, um, might like Australia, they might... You know, we brought, we bring the team out, so there's a bit of a- Yeah... get to know the culture. Um- Yeah... 43:21 and there's still Australians hanging around in various places that, you know, that might wanna join. So it's, it's just- But- Yeah... there's a lot of Americans who have worked with Australians too and enjoyed it. Yeah. 43:30 So a lot of people we've hired are like, "Oh, yeah, you know, I've worked with, for an Australian company," or, "My, you know, my partner's worked for an Australian company." 43:37 And there's, there's definitely a, there's a good feeling about Australian companies. And- Yeah, I think maybe a lack of hierarchy in the way... I think Australians can be good managers, not always, but- Yeah, yeah... 43:46 in general are very good managers. And I think that, yeah, that, we're finding that as well. People will say, yeah, they've g- had a good experience working with Australian company. 43:53 And we've got our previous networks of people we worked with, um- Yeah... and they connect, which has helped. 43:56 So, um, yeah, but it's, uh, it's not easy and I think for any Australian, any international company building into the US, it's just, you've just gotta go. It's takes longer. Yeah. Costs a bit more. 44:06 A lot more than you expect, and yeah, just hard to, just gotta really work on hiring. Yeah. The only thing I'd say is we've, um, we've put a lot of... 44:13 So we're expecting our managers as they're building their teams to spend a lot of time hiring. Um, so it's not just- Yep... HR's job to fill, you know- Yeah... recruiting's job. Yep. 44:20 It's the manager's job to, to fill those roles. Yeah, right. Which, um, yeah. They're not outbounding though, I'm guessing, but they're like- No, not outbounding. But- Just spending a lot of time interviewing... 44:29 a lot of time interviewing- Yeah... which is critical. Yeah. Yeah. Which- Yeah... yeah, it, uh, can be a lot of their time, but- Yeah... that's the way it is. 44:35 Um, so going back to the round, so you, yeah, so you were, uh, y- you're raising the round, um- Yeah... AirTree comes in. How do you think about sizing it? Um, is it, you gonna give your... 44:45 How, how much runway does that give you? Uh, was it, y- you've gone to this ads that learn is a big step up in the development effort. So w- where's the- Yeah... where's the capital going? 44:56 We wanted to give ourselves a couple of years clear runway if things didn't go well. It's kind of paradoxical in that- Mm... like, if things do go well, you almost- You run out of-... 45:05 run out of money earlier- Well, yeah, 'cause, yeah, that's right, 'cause you [laughs]... [laughs] Because you end up investing more. Yeah. So we, like, I'm a big believer in just starting from a bottoms up model- Yeah... 45:14 and just saying, well, okay, like, like, literally line by line, how many AEs are we hiring? Yeah. How many, what's the ratios of AE- AEs to CS? So based on how many AEs we would hire, 45:28 l- like how many CS people we would need to hire, how much do we need to spend on marketing to fill that quota, to- Yeah... keep them fed. Then what, yeah, what are the metrics? What revenue do they generate? 45:36 What revenues do they generate? So like a quota model, what percentage do we expect not to work out? Yeah. So you've got all, like, you sh- our spreadsheet is crazy. [laughs] Yeah. Right? 45:45 It's very detailed, but allows us to go, all right, well, this is what we, this is what we think is achievable. That's like the base case. Okay, how much would we wanna raise to give ourselves a couple of years runway? 45:57 W- and then two cases, what if things go really well? What if things don't go well? Yep. 46:02 And in each case, do we have the runway to like- Right, if things go well, you probably end up, you, you'll k- you'll essentially invest more into the team- You'll invest more... and then you can burn more. Yeah. 46:12 And potentially run out of cash faster. Yeah. Because, and, and what happens with VCs is as things start to go well, they say, "Why aren't you growing faster?" [laughs] Yeah. 46:20 You know, like, "You're growing fast, but you could be growing even faster." Yeah. Um, and they're like, "Why don't you throw more money at hiring people?- Yeah... right? 46:27 Hiring salespeople, and more money at lead generation." Um- Well, I think also expectations, well, don't think, I know, expectations have gone up on growth rates. Oh. 46:35 It's not, it's no longer triple, triple, double, double. It's, uh- Well, so we sat down with a Silicon Valley VC 46:41 mid last year, and he said, "Hey, if in your first 12 months you do X dollars of revenue," I won't say how much, um, uh, "in your first year, you will, like, be crushing it. You'll be, like, one of the best companies." 46:54 Um, I think we did four times the figure he said. Yeah. And we didn't feel like we were crushing it. [laughs] Like- Yeah... uh, I mean, I think we were, but, like, ju- that's just how much the expectation changed. Yeah. 47:08 Right? It was- Yeah... crazy. Yeah. Well, and people, yeah, 10Xs on, you know, from one year to the next and just, or more in some cases. Like- Yeah... 100X in some cases. I mean, just- Yeah... 47:18 the speed at which companies can grow once they inflect and... 47:22 But they have, you know, you go back to building a model, so, um, we are, you know, we always look at our model for Firmble and how we're, how things will build out over time. 47:29 It's just so hard to know when you're gonna get, you can't predict an inflection point. Um- No, you can't. You can't. You can't. Yeah. Really tricky to model that. And, you know, I think the better you do, the more... 47:38 Like I, I really try to push us to, to think bigger. Yeah. I think, it's probably an Aus- problem Australians have, I think a little bit, like we, we probably think a bit smaller than- Yeah... Americans do. 47:49 And when, whenever you go to the States and you hear the kind of round sizes and valuations they do over there, you think, "Oh, okay." 47:56 Like, it's not cheeky to ask for this much, or to raise this much money, to have this ambition. You should have that level- Yeah... of ambition. I think Australians just naturally, 48:04 um, probably because of our environment, think a bit smaller. And I think Aussies- Yeah... have to think bigger. Yeah. So- Completely agree. Yeah... I think you've just gotta, like, 48:13 if you wanna do 10 million in your first year, well, why, why not go for 30 or 40? Mm. Like, why not? Yeah. Like, if you can do 10, you can, you could do 30 or 40. You just have to have a different plan. Yeah. 48:23 And so, like, go for that, and if you don't achieve it, if you get- But you get a good part of the way there, then [laughs] Yeah... you've already overachieved. 48:30 It's easier just thinking through it, and it's easier probably in the early days of a startup, but just this, I've had a view that I'd like to set the most aggressive target I can reasonably, or may, or unreasonably set, maybe is a better way to put it. 48:41 Um- Yeah... but, yeah, if you only get, if you go for, pick a number, 10X or 5X and you, uh, and you get to 4X or whatever the- Yeah... whatever the- Yeah... 48:49 the numbers are, it's setting a high goal will always, uh, yeah, drive you up. What's that terrible saying? If you shoot for the stars, you could, uh- You might hit the moon or something. 48:59 Yeah, you might hit the, land on the moon or something. [laughs] Yeah. It's like, I don't know. I can't even remember the saying. [laughs] But yeah, like, it makes sense, right? 49:04 And I think Aussies need to do that more, 'cause that's what every American founder is doing. [laughs] Yeah. Right? Yeah. Like, "We're gonna do this." Yeah. And if they land here, they've still done well. Yeah. Yeah. 49:13 Good. It's harder as the business gets bigger. I remember at ACNH when, once we got into a more steady state growth- Yeah... sort of settled down, you know, many years in, went down to 30, 40%, um- It's hard to justify. 49:23 Yeah. Y- and- Yeah... you've got different style of investor as you get bigger, so it's more of a- That's true... growth equity investor that- Yeah... 49:30 uh, you know, if you missed your budgets, um, so setting too high a target. Yeah. Another thing we found was the sales team. 49:36 It's funny, once we went, we sort of, um, at one point we sort of, it was a tough year, and we sort of, we, we slowed down a lot, but once we'd slowed down, it was hard to re-accelerate, um- Yeah. Yeah... 49:45 just because everybody's, every salesperson had in their head what you can do. What- So when we were doing- Mm-hmm... I don't know, 3X a year- Yeah, true... 4X a year, people didn't mind if we set a 3X target. Yeah. 49:55 Once we'd settled to, say, 50% growth, trying to set a 2X target was- Yeah. Yeah. And it's so demoralizing for the team when you set targets that aren't achievable. Yeah. That's the challenge, yeah. Yeah. 50:06 So definitely my, the advice I'm giving here is more at the early stage, it's more like- Well, I know, I would say the faster you, while you can grow fast- Shoot for the moon, yeah... 50:12 just grow as fast as you possibly can. Just grow as fast as you can, yeah. Yeah. And, um, and, y- because, yeah, once starts th- once it starts to slow, you never, it's very hard to get it back. Yeah. Um, yeah. 50:21 And I think, like, startups a- h- are kind of like a series of, like, good problems to have. Like, [laughs] you, you wanna, you wanna create problems that are- Yeah... that, like, you're happy that you're handling them. 50:33 So to start with, it's like, "I've just gotta get a lot of top line revenue growth." Yeah. And, like, once you get that in place, then the, the worry becomes churn. Like, "Okay- Yeah... what does retention look like?" 50:45 That's a good problem to have. 50:46 And I remember at the start of the year in January saying to the team, "Hey, I hope churn and retention is a problem that we're gonna be sitting here tearing our hair out over in six months from now." 50:59 And then we had really great top line growth, and then six months later- Sure enough... we were sitting there going, "Oh, sh- crap, what about churn?" [laughs] Yeah. Like, "If there's customers leaving, what's going on?" 51:07 And then we were, and then I was like, "This is what we wanted, guys." Yeah. Like, this is actually what you want. And then beyond that, once you get churn under control, then it's like, oh, shit, now gross margin. Yeah. 51:17 Like, that, that's now your next good problem to have. Yeah. Like, are we growing efficiently? Yeah. CAC, CAC payback periods, like, these are all a series of good problems to have. Yeah. 51:26 You've just gotta tackle them one at a time, and I think as founders and early stage teams, you gotta not, like, if you look at it all and let the totality of it hit you, like, you can sink under the weight of it. 51:35 Just think about the next thing. Yeah. Yeah. [laughs] Yeah. So Sam, we do some questions that we ask everybody. All right. Um, and that's generally the worst day in the office- Yeah... 51:46 which I think we did with a Cloud Guru, but- I think we did, yeah... you've been running along enough Cutable now that you must have a few bad days. So what was a, a- There was a-... 51:53 worst, second worst, third worst day in the office? There was a- I know there's all those, but... [laughs] There was a day, I think, 52:01 like, earlier this year where it felt like things had been going really well and everyone was feeling good, and then suddenly we just had this couple of weeks where we just couldn't sell anything. 52:11 Just people stopped buying, and we couldn't work out why. 52:15 And I think in the second week of this, we were just struggling to get people to buy the product, this was, like, back in March, and then we had a, a whole bunch of customers cancel at the same time, and we actually went backwards in revenue- Yeah... 52:26 that week by quite a significant amount. And it was just, it had felt like we'd just hit this big milestone- You started starting to win and then, yeah... 52:33 everything was starting to work, and then we just had this huge setback. Yeah. 52:36 And it all happened in the same week, and everyone, I think, had been, we'd all been going for, like, a year and a half at this point- Yeah... 52:41 on this new product, and everyone was tired, and then to just be, it felt like we'd all- You, you hit the brick wall... got punched in the face. Yeah. [laughs] Yeah. 52:48 But off the back of that, we, we said, "Hey, let's just analyze what's going on here," and we stepped back and we looked at the data and we pulled it apart, and we realized we weren't onboarding our customers well, and we built this whole new 30-day onboarding program, and that has dri- that then drove, um, so much of the success of the business for the next six months. 53:07 Yeah. Um, just changing that onboarding program. Yeah. You can see all, like, the metrics show really clearly how everything improved from that program. Yeah. So- Yeah. 53:14 Just these days, little blips though to- Just needed to take a step back... and one thing I say to- Yeah... founders who are listening, just it's never a smooth line. No, no. 53:22 Yeah, we o- we, we have these nice curves of how growth looks, but- Yeah... and when there's little blips, when things go wrong, that's often the point where you learn. Actually, I can- They just, yeah... 53:30 probably the most dangerous time is when everything's going perfectly, because- You can get a bit lazy... get a bit lazy- Yeah... and just go, "I don't want to touch anything"- Yeah... 53:37 and maybe s- stop experimenting, and- It's so true... not that you want to always go bad, but when there's those little blips, it often highlights something that needs to change. 53:43 And that's where you've gotta have a culture that encourages people to be honest- Yeah... and, like, be forthright with each other, and say, "This isn't working, and let's look at the data and let's come up with a plan." 53:52 Yeah. Yeah. Agree. Yeah, that's great. Yeah. Uh, Black Friday is gonna be a big day for your customers. Yeah. It's- Massive one... a couple of weeks away. Yeah, a couple of weeks away. Our customers make 54:03 2.7 times more ads over Black Friday than they do for the rest of the year. So that period of Black Friday, they're publishing at- So people gearing up now, they're already getting- Yeah, yeah. 54:13 So they're already crea- generally- So have, have you seen volume step up as- Yeah... for example?... of production. Yeah. But they don't publish it often- So they're gearing, yeah, that's right... until November. Yeah. 54:23 Um, a lot of companies now start Black Friday sales in October, November even- Yeah... and run them for longer periods of time, but production ramps up a lot. 54:32 Um, so the weird thing- And that probably runs through, uh, into Christmas as well for people. So, so- Yeah, it does... not just Black Friday... that whole period. 54:37 The weird thing is that when Black Friday's actually on, it's very hard for us to, um, talk to customers. It's just- You can't sell when they're so busy. Yeah... 54:44 you can't sell, and even just, like, getting them is difficult. Yeah. Um, because they're just so- So you got this busy period, you're probably... Yeah, so you're gonna have a- Yeah... 54:52 you're getting people on ahead of it. Yes. And then you've gotta just accept the- And then they're just going for it. Yeah. 54:56 But yeah, the volume is the big thing they need for Black F- Friday, so we obviously help customers with that a lot. Yeah. 55:02 Um, so it's more of a, for us, it's almost more a when, the actual week that it's on, it's more of a support period. And your customers, so you said they produce what, 2.7 times more than- Yeah... 55:10 but with you guys, well, with Cuttable what, maybe even more than that in terms of the- Yeah, well, the, sorry, that's what they do on Cuttable. So they- Ah... 55:17 from, they, we, that's our stats- Oh, that's what you saw last year. Yeah, yeah... from last, from last Black Friday. Yeah. Um, they, they're producing, yeah, almost three times as many ads. 55:24 Which could be more this year. Yeah, could be even more. Yeah. Yeah. So if they're producing a lot of ads with us regularly, they're producing a lot, a lot of ads. [laughs] Yeah. [laughs] A lot more ads- Yeah... 55:33 during Black Friday. Well, the other thing we've done as agents do more work, and we've seen this with Firmable, just the amount, the ability to consume data, the- Yeah... the, we're just seeing incredible growth in 55:44 starting with our MCP server, just Claude connecting in- Yeah... and doing stuff for people, um, but the agent doing the work, that- Yeah... the ability to drive usage, um, way more than personally. Through MCP? 55:54 Yeah, yeah. Yeah. Just takes off. Yeah. It's taking off, and it has taken off. Um, but we own- So that's been using Firmable through- Through the MCP server. Yeah. 56:00 So we, our volumes went up 20 times in two months, um, through the MCP server. Um- So we're only just launching our MCP- Yeah... 56:07 in the next couple of weeks, so that's, we've, we've been trialing with customers, so it's gonna be interesting to see what that does- Yeah... to our lives. So- Well, I think just agents can consume data- Yeah... 56:15 in, in our case, and I'm, I think it'll be, well, I don't know your business really- Yeah... but just the ability for agents to drive more- Yeah... usage. Makes sense. Makes sense. Um. Yeah. Yeah. Yeah. So. 56:24 So- Very good. Yeah. Uh, and then lastly, you, you talk, it's, perhaps it's a bit of a question as an Australian founder who's built businesses globally. Hmm. 56:33 How you build a performance culture, I think as Aussies, we like being great blokes, um, great people. Yeah. Uh, how do we... And we like, I think we're very focused on building great cultures. Yeah. 56:44 I think one thing we struggle with at 8connect, I, and we're just, you know, with Firmable, I think we build a great culture too, but a, that performance culture, how you build that? 56:52 I don't- Or how you blend that with it. I, the, the things I always look for in staff is what I say, I call curiosity and drive. Mm. I think they're the two things I want more than anything. 57:03 So are they curious enough to ask why? Mm. To ask why is something happening, and then do they have the drive to go and improve it, to go and implement it, and to, to push it through? Um, we had this, uh, 57:17 value, uh, one of the values of Cloud Guru, which I absolutely loved, championed, and we, we have a very similar one at Cuttable, is thoughtful but decisive. Yes. 57:27 Um, so I think you need to, it's kinda what I've been talking about this whole podcast, is like, you need to be thoughtful enough, have a plan, collect data, but then you need to act. Mm. 57:37 And people, I think, often aren't decisive enough. Um, uh, so yeah, I'm, I, I think you've gotta have that constant blend where you're moving back and forth between 57:49 capture enough data to make a decision, but then actually make a decision, act, and learn from that, knowing that your decision isn't gonna be perfect. Yeah. 57:57 And so I'm probably not gonna say, "Oh, you just need to, like, whip everyone," like [laughs] like some other co-founders or, sorry, founders might say- Yeah... especially in, in America. 58:06 Like, I, I don't, I'm not that kind of hard- Well, some crazy culture, man... hard-driving- Yeah. Some companies are working seven days a week, and just- Yeah... 58:13 basically saying if you're not in the office every day, you're not working here. That's it. Whereas I'm, I think I'm probably more about outcomes- Yeah... and being thoughtful about, like, 58:21 are we making the right decisions to get the outcomes, and then are we getting the outcomes? Yes. Yeah. And if we're not, like, what are we gonna do? What are we gonna do? Like, you have to act. Yeah. Um, yeah. 58:30 You mentioned curiosity. I just think that's, just in, in the, in the, the world changing so fast- Hmm... curiosity, not just as a cultural value, but just as a inherent behavior of everybody these days. 58:41 That's what I say to my kids. Yeah. Like, you know, when they talk about what am I, what should I do, or like- Yeah... what do I do when I finish high school? 58:48 And I'm like, "Well, I think the best thing you can do is just be really curious and stay up- Yeah... with what's happening." And learn, yeah. Yeah, like with AI, just use it- Yeah... 58:55 all the time, 'cause whatever you're gonna go into, you're gonna need it. Yeah. Yeah. Great. And one final quick question, advice for somebody starting today? Uh, one, solve one problem at a time. 59:07 You can't- Can't solve the world... it feels like if you, the, if you're looking at the mountain, it feels like it's gonna fall on you, you know? [laughs] Yeah. Like, building a business is really hard- Yeah... 59:17 and it takes a long time. There's a lot of things that'll go wrong. So just, like, solve the next most important problem. Always focus on that, then go on to the next thing. Yeah. Great. Awesome. Thanks, Sam. 59:27 Great to hear about the journey, and- Thanks, Lee. It was fun... glad to have you in the, uh, in the guest seat. In the hot seat. I'm gonna do this for you soon, for Firmable. [laughs] Yes. Yeah. Happy to do it. 59:34 [laughs] Not right yet. Yeah. But, uh, well, sometime, but, um, no, but it's been great to hear the Cuttable story, and, uh, yeah. Oh, thanks, mate. Awesome. It's been great. Yeah. Thanks. [outro music]