How FLORA is Reinventing Creative Design with AI

None of the companies I'd seen I would have wanted to start myself because they weren't creatively interesting to me. The second thing I realized is that I wouldn't back myself because I wasn't a one-on-one founder. And the third thing I realized is that I really miss art and I wanted to explore that [music] as part of my quarter life crisis. The reason why creatives don't make money is because their tools suck. You know, when I was in investment banking, I could model out an entire business model in 2 days. We would sell the company and we would make 10 million. The reason why I could do it in two days is because I had this software program called Excel.

Meanwhile, I'm at ITP learning 10 different creative tools. [music] Takes weeks for me to to get the creative output that I want. We view ourselves like the Nike of creative touring. Like Nike was founded by Phil Knight, a runner, to build shoes for himself and for other runners that he used to run with. It's a company founded by athletes to enable athletic excellence. [music] Explorer was founded by creatives to enable creative excellence. Welcome to Subversive, a podcast dedicated to sharing stories from the best consumer subscription apps in the world. We'll bring you lessons for how to grow your consumer subscription business, [music] including insights and inflection points that led to exponential growth from leaders at category defining companies and [music] innovative startups.

Let's get into the show. This episode is brought to you by RevenueCat. The world's best apps use RevenueCat to power purchases, manage customer data, and grow revenue on iOS, Android, and the web. Join more than 100,000 apps that are already using RevenueCat to grow their businesses and track more than 10 billion and spend, including category leaders like OpenAI, Notion, and Ladder. Learn more at revenuecat.com/subversive.

My guest today is Weber Wong, the founder and CEO of Flora, an AI powered platform that helps creative professionals ideate, iterate, and scale their ideas faster than ever before. In this episode, I talked to Weber about how his experiences as an artist inspired him to found Flora and influenced his early design thinking, about the steps that Flora has taken to establish product market fit with a highly engaged group of early adopters across the brand marketing, CPG, and fashion industries, and about how Weber expects AI to continue reinventing the market for creative tools in the years to come. All right, welcome Weber to the Subversive podcast. I'm really excited to have you on today to talk about your startup Flora. And before we jump into your background and what led you to start this company, let's just start

with what is Flora and how would you describe it to listeners who maybe haven't heard of the company yet? Flora is a full creative operating system with every single model across text, image, and video on one infinite canvas. You can connect these models together to concept really quickly through your creative ideas and build scalable generative workflows um for repeatable creative processes. We're focused on serving the world's best creative teams um including folks like Pentagram, Nike, Lionsgate, folks like that. We sort of view ourselves as like, you know, Harvey, but you know it for law for the best creatives in the world. Um and yeah, we're trying to build a really a really great power tool for creatives.

Awesome. And what sort of creatives are using the tool already? I imagine it's a fairly diverse group, but is it, you know, brand creatives who are building ad campaigns? Is it artists, architects, designers? Is it a little bit of all of the above? It it is literally all of the above. And um I think that's because all digital media is essentially a combination of text, image, and and moving image, which is video. And those are the three fundamental building blocks for text, image, and video. By connecting those together, you can access every single creative model like text to image, image to video, and things like that. So, we've seen adoption from all these different industries of folks building very specific workflows around their creative process. If I were to kind of roughly rate um you know, the magnitude across different industries, I would say brand and marketing is definitely um you know, the

biggest use case. uh fashion, CPG, shoe design, things like that is I would say second most. Um and then underneath that would be I would say like film, VX, VFX and previsualization. Uh so that'd be folks like Netflix, Lionsgate using it to quickly prototype scenes. And then there's like a bunch of, you know, small business owners using it too, like this guy who owns a a tire shop out of Montana that I talked to who's using Florida to make ads for his own tire shop. or my founder friends cloning their competitor's ads and you know I really enjoying that or using it to make their own you know landing pages. Um so it's been fun to see the different ways in which it's being used.

>> Yeah. Well that's such an encouraging sign for a startup that's only been around a couple years to see so many diverse groups of people using it. So excited to come back to that in a bit but before we do I want to talk about your background. So you have a very

diverse set of professional experiences. You've had stances as an investment banker, geopolitical researcher, a venture capitalist, an artist, and so what inspired you to try so many different paths earlier in your career, and how has that prepared you to build this company? >> I mean, I I'm fortunate and cursed in the fact that I find a lot of things interesting and I see a lot of parallels between things. So, my two biggest interests initially were poetry, just cuz I don't know, it's poetry. I just loved it. Uh, and history. And I loved history so much that I started studying evolutionary history which is what I was doing at Stanford to go you know even further back and venture capital is in many ways a similar study of the delta across time but applied to the future.

Um so I was able to satisfy that urge when I was working at my love interest. But the general trajectory to that path is like really wanted to do poetry but realized that I needed to make money for a living. So I went into investment banking and uh the thinking there is work at a top investment bank, get a job at a top venture firm, invest in the best startups and then learn how to do my own. So I worked at this firm ventures for a few years and that was really fun. Looked at everything from you know global supply chain software to a lot of the AI stuff in the early days.

Um remember looked at like anthropic runway folks like that. Um after my time in venture after a couple years there I realized a few things. One is that um none of the companies I'd seen I would have wanted to start myself uh because they weren't creatively interesting to me. The second thing I realized is that I wouldn't back myself cuz I wasn't a one of one founder. Um I didn't live up to my own standard. And um I didn't have a unique worldview or a hill that I would die on. And the third thing I realized is that I really missed art and I wanted to explore that as part of my quarter life crisis. So, I decided to leave my job, move to New York, work at a coffee shop, and get into get into the conceptual art world. Uh, just by chance, the week before I left, I looked at investing in Runway ML. And, um, that was interesting because it was the first company I'd seen that I might have wanted to start myself. So, I kind of

asked the founders what the story was, and they were like, "Oh, you know, we're from NYUITP where we use technology to make art." And I I thought that was really interesting because um, you know, I didn't know you could do that. I thought you could only use technology to make money. So when I got out to New York, I got into the Brooklyn Art and Taxi and met the head of ITP. Um, and we really hit it off. He saw a lot of potential in me. So he gave me a full scholarship and let me join ITP right away. I went from being an enterprise software investor to just making art. Uh, so I learned how to code, learn how to do installation art, some physical computing, and I started working on these like real time AI paintings. It was like a webcam. would take real world data, cut it up 20 frames a second, run it through a hyper fast image to image model and then display the result in 50% strength. So

you could see yourself in a Picasso painting or whatever, you know, [clears throat] prompt you wanted that started getting a lot of traction in the art and tech world and people started asking me how I did it. So I built a tool for other people to be able to do that and that was the first version of Flora. It was a nodebased tool. It was webcam, real-time AI and output. >> And then people started using it a lot and it started costing me a lot of money. Um, so I thought I need to do something about that. My initial idea was to start an agency and build a fully autonomous creative stack. So a single creative could do the work of 100 because one thing I realized was that um creative means of production are broken.

The reason why creatives don't make money is because their tools suck. you know, when I was in investment banking, you know, I could model out an entire business model in two days. We would sell the company and we would make 10 million and uh my managing director would take all the credit. Um, and the reason why I could do it in two days is because I had this software program called Excel. I didn't have to calculate every single number.

Meanwhile, I'm at ITP learning 10 different creative tools. Takes weeks for me to to get the creative output that I want. They're extremely clunky and um really not that well-designed. Um so the idea behind the agency initially was to essentially create the Goldman Sachs of creative services. Uh build this really powerful tool, go find the best creators in the world, pay them a bunch, and have them do amazing creative work. And then I realized that I don't know anything about starting an agency.

Um I don't have any kind of customer connections or anything there. Instead, I should just go build the world's most powerful creative tool, sell it to every other creative team, scale like a ventureback startup, and down the line, we can do creative service if we want. And today, a lot of our customers have to collaborate with us. And if we were an agency today, we would make a lot of money. Um, but the reason why we didn't at first is because the tool we're going to build is going to do so much of what used to be wrote manual labor in creative teams that it's unclear what the agency itself will look like in a couple years. And from a self-preservation standpoint, it's always better to be on the side enacting the change than reacting to it.

And from a creative technologist side, I learned that if you really care about your creative output, you build your own creative tool. And I love building creative tools. It's really fun. Um, and most importantly, I realized that I would back myself for this. There's a clear, very large market opportunity that no one else sees because most of Silicon Valley are not creatives. Um, they don't even know the creative professional industry exists, much less how it works. And we fundamentally care about enabling great creative work, not slo. So, we wanted to build that power tool. Um, and that's when we decided to start Flora. That was about two years ago now. We launched the product about nine months ago. And yeah, we've been scaling rapidly ever since.

>> Yeah. Well, there are a couple things you said there that I find really refreshing. One, I had a previous guest on the podcast, Max Marcy. one of the co-founders of Superpower and he talks about how American culture tends to embrace these linear stories about how people build their careers but in reality some of the most exceptional people have very nonlinear career paths and I think that's because some of the the biggest innovations happen when you have these intersections between two worlds so in your case it' be the intersection between technology and art uh and so I think this meandering path you took between investment banking in venture capital and art and now being a tech founder I think perfectly sets you up to be building this company. Um, and then two, I loved how you talked about the importance of being the one creating the disruption rather than reacting to the disruption. And I think that's never

been more important than now given how rapidly things are evolving with AI. Which brings me to my next question and

that is you started this company 2 years ago. You launched the product 9 months ago. I imagine even just in those nine months a lot has changed. So what did the product look like in its earliest form and how has it evolved over those last 9 months or so? >> So before we launched the full product, we had an alpha version and interestingly that alpha version was built on top of that first version of Flora I made for my art school and agency friends. Um and that initial version had a node for every single model. So there was like a node for like a text to image model, an image to video model, every single school. We realized that it wasn't a generational product because it was too complex. It was too technical for too many people. So when we launched the full version of the product, we just scrapped the alpha and rebuilt them from scratch in a new abstraction. Just text, image, and video. If you connect text to image, we

get all the text to image models. If you connect image to video, we get all the image to video models because creative professionals do not care about Hyo Miniats O2. They care about having a great video. So we wanted to kind of remove that abstraction from them. And that also is a better building block for the control and agentic features that we're going to build on top of it. The current version of floor is just I would say phase one of three. Yeah. And in terms of you know what we've learned it's been pretty crazy because like when we launched um there were people that use AI creative models and then there were professionals and there was no overlap. Um so when we launched the product we were like hey guys like current AI creative tools are built by non creatives for other non-creatives to feel creatives. Um let's go build an AI creative tool for professionals. In hindsight it's very obvious. Um, we

launched with that premise and it immediately went viral among professionals because they were like, "Wow, someone's finally building a tool for us. It would be great." [laughter] Um, and we've started to see that ven diagram start to like emerge in that middle portion of high craft professionals that are also getting more and more AI lit as well. And that's kind of how we view the market across these two axes. Um, those people are cracked because they have like the taste they need to make something great. But by building these kind of um AI workflows, they're essentially quantifying their creative process. Um and we've seen emerging behaviors of new types of creative teams emerging um that are doing things like like typically the the brand identity process um costs, you know, hundreds of thousands of dollars, takes six months. Some of the, you know, creative professionals that are very AI

forward are offering a new creative service called a generative brand script. And they do that entire process in two weeks at a far lower cost. And it's doing extremely well cuz the result is still the same. It's about storytelling, um, communicating your vision, things like that. It turns out, you know, you don't need a bunch of rope manual labor to do it. You just need good concepts. Um, we're seeing folks like Lionsgate generate entire content, movies in Flora, taking like 1,000 story beats and tests, generating 1,000 story boards, generating 1,000 videos, stitching those together and putting them in front of an audience and having them rated on a scale of 0 to 1200. It was high enough to go make the entire film. We're seeing people instead of handing off like static brand books, hand off workflows that are essentially onbrand asset generators and codifying the creative process. In general, we're

seeing an emergence of a new type of creative team that is not built around wrote manual labor, but around creative systems um that let them kind of scale their ideas and uh focus on their good ideas. So something implicit in what you just said is this contrarian approach

you've taken where unlike a lot of other first generation AI tools that are focused on single prompts or outputs think like midjourney you've you've taken this platform approach where you're focused on a workflow going from yeah you start with a single text image or video input but then you just iterate and iterate and iterate and you have this infinite canvas. What was the moment when you realized that was the right form factor for for these creative professionals and how has that influenced other parts of your product vision your go to market strategy >> before I even started Flora? So at ITT it's a program where we use technology to make art. Uh it's a rare collection of folks that are technical but highly creative. Um it is very common there to use nodebased workflow oriented tools to build creative systems that are reactive right there's tool like cost designer blender nodes things like that right um

I think it was very obvious that if you have a model now that can make a piece of media like that um let's not build a creative tool around making one piece of media that can be done like that now the creative opportunities to take a step back, connect all these models together and design the entire creative workflow. That's the ultimate creative power. And um the reason why a new interface is required now is because we're in a new computing paradigm. So Adobe, you know, they've been around for four decades. Um they were founded to build the creative interface on top of a personal computing paradigm where it was about controlling every single pixel on the screen. Um, since 1987 when Adobe Illustrator came out, all image editing has been about altering and reordering layers. That's it. And Figma is just a WebGL translation of that. Since Adobe Premiere in 1991, all video editing has been about altering and reordering time,

right? Um, traditional creative tools are bottoms up. Piece by piece, layer by layer, you kind of build up your creative output. It's kind of like crafting. And it's about making one piece of media at once. Creating with AI is the opposite. It's it's tops down. You just ask for what you want, you get a rough version, and then you ch away from there to get to pixel perfection. It's kind of like carving, right? Um so that's a net new creative process and that definitionally requires a new interface. One, because it's a different direction of creation, and two because of what I was saying earlier of like you could do the entire workflow now. Um, so as a result, it was very obvious from day one that that was the right approach for the future because it just was way more powerful for creatives. It's not an easy interface to design and build for sure. Uh, but there's really no other choice in our view for how to build the

most powerful creative tool possible. >> Yeah, I loved how you talked about it in terms of bottoms up versus top down creation. And I guess the analogy for me, and I'm no art expert. I think I took art 101 and 102 in college, but beyond that, I have very little art background, but I know Michelangelo famously talked about how he could see this the finished sculpture in the block of clay. And so, as you're describing working with AI, it's sort of like, okay, you already have this vision of what you want to create >> and now AI is helping you to sort of peel back layer by layer, but get to a much more brilliant output much faster than you would be able to if you were building everything from the bottom up.

And so it sounds like this canvas multi-step creative flow approach to Flora, you already had a lot of reasons to believe it was the right approach given your experience at ITP and working with these other people at the intersection of technology and and creative design. When did you see that validated by your early adopters? Like at what point did you know you had the initial signs of product market fit and how did that show up in either your metrics or in qualitative feedback you were getting from your customers? I think there's different levels of product market fit. From day one, people thought the concept was cool and interesting, even if they didn't know why. Late last year, there were folks from Pentagram reaching out and they were using Florida to generate hundreds of logo variations instead of doing those by hand, which would take days.

That way, they could just see them all out at once, choose one and then refine that, right? So, they're early signs of, you know, really interesting use cases. I would say after we launched for a while, you know, people were using it to explore really quickly the concept really quickly. But as people started to understand how to use this tool more and as we started to understand the power of the tool more, as people curated it to their specific creative process, we started to see more of this usage of Laura as a creative system um of like them codifying what they used to do by hand so they could move way faster.

entire brand identity workflows, generating constant movies, um things like that. And once people started doing it like that and we started educating people more in that kind of uh mode of thinking, which is more of the future mode of thinking, um the average nose generated per user for those folks just skyrocketed. I would say that was true product market fit. And then you know the smile and retention curves and stuff like that. That's when that started to come in. A lot of the world is not there yet. I would say like literally there are probably 10,000 people in the world or teams in the world that actually think that way, but the ones that do are doing very well. Um, so that's been really interesting to see. Got it. And I love how you talked about different levels of product market fit there. I think often times as tech professionals, we are the opposite. It's sort of like, you know, product market fit when you

see it because there's just overwhelming evidence that all the metrics are moving up and to the right. But the reality is if you're an early stage startup, you only have so many early customers, which means there's only so much quantitative data you can use to validate product market fit. It sounds like some of the first indications you got were just people at some of the top design agencies in the world telling you, "This product's really cool. We want to be part of the alpha. We want to get our hands dirty because we haven't seen anything like this before." And then you started to see that manifest in the smiling retention curve or in the growth metrics. Uh I think this is a good segue into the next topic which is just around how crowded a lot of these AI tools markets have become very quickly because it's this exciting period in history where there's so much innovation happening there's so much opportunity to

be had so much money to be made. Um

what were some of the early design decisions you made to avoid just being another AI rapper? And I think you've already talked about this to a degree, which is you're taking this fundamentally different approach, but I guess I guess the more general question is like how did you break out above all the noise? Because there's so many AI creative tools out there and and so what were some of the other ways that you you stood out early on?

>> I would we never like thought like okay like what are people not doing? Let's go do that. We thought about like what do I literally need? Like what do creatives literally need? What is right? What is the creative interest going to be in 5 to 10 years? And we just started building that. that turned out to be right. Um so and the only reason why I started the startup is because it was very obvious that no one understood that. People barely even believed us at first. Um and that's when you know that like you have a pretty contrarian and non-consensus view and if you also happen to be right you know that's how you break out of the noise right. Um I think the world is starting to come over to our point of view both customers and other kind of creative tooling companies when you see a lot of success that that tends to happen. Um, but you know, at the moment I'm really not that interested in

like I I think for our product category, we're like year one into this product category that we've created out of at least 10 or 15. Like we have so much more to build to turn it from like 1,000 people using it as a creative system and you know hundreds of thousands of people using Flora um in a less deep way. We need to get those hundreds of thousands of people using floor in a deeper way. And also, there's the entire, you know, hundred billion dollar creative industry that is still just getting their feet wet and has not even generated an image yet.

>> Um, so there's definitely a lot of room to to go from here, I think. Um, and another way that we cut through the noise, which isn't intentional, it's just who we are, is um, we're the only tool in the space that's actually built by creatives and it shows, right? Um, like if you look at a space like cursor, yeah, like everyone that's building that is also a coder. Um, in our space that is not the case. It's like random people pivoting from other things. Um, >> and I would say that like if we are we view ourselves as like the Nike of creative touring. Like Nike was founded by Phil Knight, a runner um to build shoes for himself and for other runners that he used to run with. It's a company founded by athletes um to enable athletic excellence.

Flora was founded by creatives to enable creative excellence. That is a multi-deade, you know, endeavor and that shows through in the product, in the brand, um, in the people. And I think that's also a big differentiator for us as well. >> This episode is brought to you by Paddle. Today's fastest growing apps are multi-channel. They use web monetization to reach new audiences, improve attribution, and boost margins. But selling on the web gets complicated fast. Managing currencies, taxes, fraud, and disputes gets increasingly difficult as companies scale. Paddle solves this.

As the merchant of record for thousands of apps, including Runna, Goodnotes, Codeway, and HubX. They take on the complexity of web monetization. so you can focus on growing your business. Learn more at paddle.com/subversive. Yeah, I love that. And and to me it brings us full circle to one of the things you started with, which was you were at Menllo Ventures as sort of a junior venture capitalist. You were looking at all these startups and these founders and you realize like I wouldn't back myself. I don't even meet my own standard. And I think the reality is there are a lot of you know former product managers, marketers, VCs, investors who decide I want to start a company and they go try to build a product for a target market that they sort of view top down as this is a big market, there's a lot of money to be made, but they're not deeply authentic to the problem they're trying to solve.

And what you end up with is sort of a derivative of we're going to build an AI powered version of Figma or we're going to build Figma for professionals versus no, I'm going to go to ITP and I'm going to figure out what it looks like to use technology to create art and then I'm going to bottoms up understand the problem that needs to be solved and that will guide all of my product decisions and all of my go to market decisions.

And it sounds like that's the path. It's the much harder path, but it's the path that you've been on for these last few

years. Let's talk about your early adopters and your ideal customer profile because you talked about at the beginning of the podcast how you have all sorts of different creatives using this tool. But are there a particular pocket or set of pockets of users that have most resonated with this product in the early days >> before starting the company and you know early on it was really fostered out of this kind of really niche creative technology scene. We were doing like real-time AI to enable creative technology projects that do not make any money, right? We're just trying to make cool stuff. One time we were talking to a creative tech agency and they were like, "Hey, if you guys really want want to make money, stop serving us.

Honestly, like you guys should go focus on other folks." Um, and when we started moving towards more like asynchronous stuff, like you know, the higher quality image models and video models, um, which is still much faster, although not real time, much faster than how they currently do things. Um, that's when we started to get kind of industry interest from a lot of these folks. I think early adopters of the alpha, you know, were a wide range of folks. Um, folks from Pentagram, folks from a few other kind of top creative teams, but at the time the only people really experimenting with AI were like AI consumers on Twitter saying like, "Oh, look at this AI film I made. This is going to kill Hollywood and it's like kind of bad, right?"

>> And like there were a lot of those people kind of trying things out. Um, but you know, we we knew that that was not as much what we needed to focus on. after we launched I mean we got interest from everywhere like from obviously like brand and marketing agencies in-house teams uh fashion CPG is fast becoming uh very very big for us even though we didn't expect it because in addition to visualizing the brand and marketing you know we're helping Nike visualize the shoe itself so it's kind of like full stack in many ways but also like architects seem to love it a lot as well so at first actually there were too many different vertles in which we are seeing interest and at the time that was fine for us like we're creative technologist where um almost anti-use case focused from that perspective and we have a general philosophy that you know all media and any creative tool um should be able to be made in Vora if everything's

made out of text pixels and moving pixels. Um on the GPM side though as we matured over time and seen specific adoption in different areas we have focused more on like the fashion CPG and brand and marketing verticals over time um for a couple reasons. One is just like we're seeing the most adoption there. Um two is that we're able to get to production grade uh at a higher you know at a higher likelihood than for a few other verticals just because AI is not there yet.

>> Got it. So, brand marketing, brand design, and then fashion and CPG. Those are some of the clusters of early adopters where you've seen the most

traction. What are some of the specific growth strategies and tactics that have been most important for growing your early subscriber base within those markets? >> Launching product, we haven't done anything very sophisticated yet. Um, we've just announced features on Twitter. That's gotten quite a bit of word of mouth. Um, and I I think the word of mouth has been maybe the biggest driver. Uh, we have pretty steady kind of signup growth every week without doing too much is because we're primarily like a product focused company. I think one thing we do that's different than other folks is um we do a lot of interesting stuff on the brand side. Um, we're connected with a lot of the top creative teams in New York and uh much more plugged into culture than a lot of the other folks in our space. So we'll do stuff like host generative media discussions at the floor office. We'll invite folks from

like Pentagram, A24, like all these top creative folks in New York and discuss like what does great work look like in a generative age um and talk about it from that level. We just put out something called field guide which is an operating manual for the future creative team. Uh because a lot of teams now come to us and ask us not just how do you use Flora but how should I structure my creative team and process in the AI age. Um, so that's been really interesting as well.

Uh, we have this kind of creative partner program where a lot of folks are kind of posting workflows and sharing their work on socials. That has been great as well. Um, oh yeah, and there's something new that we're doing that has been working very well called for employee creatives. Um, one thing we saw early on is right after we launched, we would go to creative teams and they'd be very interested. They'd be like, "Oh, this is really cool. Like what's going on?" We would walk through all the functionality um and neither us or them would really know how to get the most out of a tool because it was so powerful and so new we didn't know what to do with it. Um we've now worked with hundreds of the world's best creative teams. we have maybe the best knowledge in the world on how AI tools should be used in the creative process and um it needed to scale beyond me kind of telling our customers and a lot of our

customers have tried to join us um especially the more AI experts at our different creative teams that we work with. So we brought in a few to go and procilize you know the AI native creative process to the other worldass creative teams. uh we've hired folks from Pentagram, from a lot of these other places uh in house and when we bring them to uh our customers and work with them, they do a better job than me even on showing them how to contextualize it for their creative process. Um in some ways they're they're kind of like the agency's agency and it's a really fun job just to be able to go to these great creative teams and and work with them and try to solve their problems. Um so that's been really interesting to see as well.

>> Yeah. Yeah. I mean, what I'm hearing you say, and it's funny because I'm seeing this across some of the very best category leading AI companies in many fields is right now if you are doing really amazing cutting edge product development and innovation using AI, that's almost the best thing you could be doing to grow your user base as opposed to traditional marketing because we're in this rare moment in time where so many people are so open to trying new technology that the best products just sort of sell themselves. I've seen this with gamma and AI powered presentations.

I've seen it with Perplexity. I've seen it with uh 11 Labs, Whisper Flow and Voice Dictation with um we have a client called Study Fetch that's doing this with a AI applied to high school and college education. And so that's what I hear you saying is like we're just it's almost a movement like we are reinventing how creatives run their process using AI and we are doing a lot of this in public. We're bringing along some of the other thought leaders and experts in the field. We're putting new literature out there. We're having we're hosting these talks and these gettogethers where the luminaries in the field talk about how things are changing and what you can do to take advantage of new technology. And that is its own sort of growth strategy.

>> Yeah. And I do think one thing that's slightly different about those companies you mentioned which are um easier time to value and less like depth versus tools like us, Decagon, Sierra, Harvey which are like kind of deeper applications um are that like the deeper application companies their software embeds into what used to be the creative team and it requires some sort of like change management there >> and if it's pulled off it's highly effective for the team, but it requires a little bit more kind of integration.

And um our forward deployed creative program, we got conviction behind it, you know, not because of Palunteer, that was like a little bit too far out, but like looking at folks like Harvey and Lora um having legal engineers and that playing out really well. And so far, you know, we're the first people to do this in the creative space, it's playing out really well, too. Um, so it seems to be maybe a generalized um, strategy for uh, you know, deeper application AI startups.

>> Well, and what I hear you saying, if I'm understanding correctly, is your tool is deep and complex enough that it requires some pretty fundamental behavioral change, which is hard. Changing human behavior is one of the hardest things for a tech company to do. But if you manage to do it and if you are the company that is writing the manual on how to change your behavior as a creative in this new AI era, it becomes a really powerful competitive mode because now you've written the rule book, you have all the relationships, you've got very deep product market fit.

And when inevitably the next floor competitor comes along, it's going to be really hard for them to break through that mode. And so that brings me to the last topic I wanted to cover, which is moving from early adoption to sustained usage. Let's talk a little bit more around I'm sure you've gotten a tremendous amount of feedback from all these early adopters across all these different sectors that you touch, but let's focus on brand marketing, CPG, and fashion since you said those were the segments where you're getting the most

traction. What sort of feedback have you gotten from these early adopter customers, good and bad, and and how have you learned to filter which feedback needs to be prioritized versus which can be ignored for the time being? Obviously, the the learning of like providing a little bit more education and handholding in how to get the most out of it and also providing better resources online for them to do it kind of self-s serves.

I think learning how to better communicate product value. Um, I think as someone with a background in poetry, I've realized that poetry is about expressing yourself, whereas copywriting is about expressing to someone else and translating someone else. And um, for different creative teams, that's different things. Um, so we've learned a lot about um, how to be very specific with the value prop, especially for a modular tool. This is something similar to I'm sure what notion went through as well in the early days. I think in terms of what product feedback we take, you know, we're not a very like lean startup, you know, YC, let's pivot and pivot and pivot. We've never had to pivot. Like we knew exactly what needed to be built. The product vision has been exactly the same. We're still only onethird of the way through it. I think it's on the the margins of like how it get gets executed very specifically or

I'll put out a first version and the first version we put out in February the the skeleton of that was correct but there just some small things we need to add in there that are quality of life things or emergent behaviors we noticed and if we hear multiple of that in the same week and it's very clear that they need that then we'll add that along the way while still progressing towards you know step two and step three of of what we're trying to do. Um but the the core innovations do not come from um listening to all the noise and then trying to turn that into a bunch of small features. It comes from first principles thinking of what is next.

That may not be how other folks in our space operate. Uh it's just how we operate because we know that there's a very clear kind of path here and we need to be very true to that. Well, and it goes back to why building for your own need is such a superpower and being so deeply authentic to your your product vision, your company vision is so powerful because you already kind of know intrinsically what the future should look like for this space and you can just build for that. Which means that it's still important to get customer feedback to sort of validate things on the margins as you go, but you've got your own internal northstar that you can largely follow. So that's sort of the art behind, you know, finding and building on your product market fit. Let's get to the

science. As you've become a larger company and as you're raising additional outside capital, I imagine there needs to be more numbers put behind all this. So what specific growth metrics or behaviors are you looking at now to understand whether Flora really is getting broader adoption, deeper retention, and so on and so forth? Yeah, we were really bad at all that starting out because uh I've never done a product before. I just made art. Um I made interactive art which is interesting because like you literally build it. You see how people react to it in person. So you're not like looking at post. You're just like looking at their face and then you see like whether they like it or not. You tweak the code, you run it again. And I think we're all very intuitively great at that of like that really qualitative version of product building. Looking into someone eyes or asking them how they feel. Um going

deeper than that. um understanding of deep motivations and trying to build a product vision around that. I think in terms of metrics we just uh a couple months ago brought in our head of product tric who founded and scaled a design education business called design lab uh to 50 million in revenue and he's seen so many different tools come through and educated people on these tools. So our onboarding is a lot better now because of him because he's seen a lot of those things. Um I would say the way we think about metrics is on a weekly basis and in terms of a few key kind of northstar metrics. So on a weekly basis we'll look at how many signups do we have coming in um the conversion to paid um and then the the weekly churn just to understand on like a weekly basis how we're improving whether a certain onboarding change improve things or whatnot. Uh on a northstar basis we'll look at nodes generated. We look at this instead of uh

weekly active users which we do look at but it's less important to us because of that behavior of like the people that adopt it as a creative system um have 100 to 200x more usage than the person that is using it um as a better midjourney. Um so we really index towards that metric because we feel that uh it captures that. >> Got it. So, it sounds like, as you'd expect, you've gotten more and more sophisticated around tracking the right KPIs, the right growth metrics, iterating on them from week to week, month to month, getting getting more and more robust around how you're evaluating the overall health of the business and its growth. Um, to what extent have you

started running lots of AB tests in the product? And are there any particular, you mentioned onboarding experiments, are there any particular growth experiments you can point to that were either huge wins or huge failures? We we we don't yet, but we should soon. Big experiments that worked. [laughter] Um we made some of our buttons green and uh instead of white surprisingly recently that that has been helping. Um we introduced project level sharing recently and that has increased kind of uh the number of signups we're getting through organic sources as opposed to just top ofunnel through like marketing and things like that. Uh we redid our entire onboarding at one point and that was pretty important. Um where we went from, you know, showing them a bunch of big example workflows, uh which is more the philosophy of like here's a fish. Uh and we switched to more of like a here's how you connect text to image to video

which is more like let's teach you how to fish and uh that has gone well. Uh but there's still some improvements there where ideally we can show them a big fish and motivate them and then teach them how to fish. So still some work to do there. I think honestly based on the conversation we had a couple weeks ago on growth loops, we're starting to look a lot more at stuff there. Um, and have a lot of additional things kind of coming out there and we'll start tracking that more specifically.

>> Cool. Well, that's exciting to hear. And what I love about your response there is it's everything from completely overhauling your entire onboarding flow all the way down to just the color of buttons, >> which which is generally something you don't start to optimize until you're much larger company. But I've heard all sorts of stories. One of my favorite stories related to the color of buttons is I've heard stories of startups changing their CTA buttons to be the same or very similar color gold as Amazon uses. And because so many because everybody uses Amazon Prime at this point, they literally see a lift in their conversion metrics just because consumers are familiar with Amazon Gulf and so it leads to higher clickthrough rates.

>> Uh let's get to one more question before we get to our lightning round. How do

you think AI will continue to reinvent the market for creative tools over the next three to five years? And to the extent you can share what's your product vision for Flora by 2030? >> You know, I I think the time of purely just obsessing over the model and you know doing test in to oh look at this cool output I made is starting to kind of die down. The models will continue to still get better for the foreseeable future, but now it's about how do I use this in my creative workflow? how do I use it to make great work? Um, this was the year in which professional adoption started to happen, but it's still very very early. So, I think next year it'll start to become a lot more mainstreamed.

Um, and that'll be really interesting to see, you know, how that plays out. And I think we uniquely have an advantage and an understanding of um what that will look like. It'll probably start with a few power users in the org. As we unlock more growth loops and you know better sharing and more features to enable kind of a power users knowledge to spread. Um we'll start to see a lot more of that broad usage within teams.

Um in terms of our product vision I'll put it this way. So like when the personal computer came out Microsoft was the first to build basic on top of all these different personal computers and there were a bunch of them right? But over time they kind of became the standard. Because they became the standard that gave them the right to build DOSs on top of that, then Windows on top of that, then office on top of that. For the generative computing paradigm, you have all these different models, especially in the media space.

Um, we're currently building the first layer right above that, the creative computing layer. There is so much left to build on top of that once we have that initial visual programming layer done. how that specifically evolves. Um I I'll keep it high level, but um how do you make it easy to build these visual programs that are both highly powerful? Um while making it just really easy to build it because like creatives don't know how to program. Even visual programming thinking in steps is new.

They're very focused on like laser focusing on one piece of media once, right? Um the second thing is how do I share my programs? uh if I have an interesting workflow um how do I share it with someone else on my creative team? How do I scale that? How do I let more people use those creative programs? And um there's not um there's not too many or there's there's no direct examples of how we should do this. um we're taking a bunch of metaphors from a bunch of different spaces from literal programming to you know things like nadn to you know document based workflow things um but you know those are the questions that really have to be figured out and there's a lot of different layers on top and interface interface yet to build for sure um but yeah as well as just in general um we want to have no sacrifice of control relative ative to traditional creative tools. Um, I get frustrated with AI a lot myself. I I've used 10

different traditional creative tools. I love the feeling of being in complete control. Um, AI doesn't provide that yet and we're going to layer that in over time as well. Well, it sounds like a really compelling vision and because you're so authentic to the company and product you're building, uh, I love that you already have a very clear five-year roadmap laid out and as a potential user, I'm I'm excited to, uh, take advantage of whatever tools you build. So, let's get

to our lightning round. And I always like to start with the same first question, which is, what is the best part about working at Flora? And what's one thing you would change? Or since you're the founder, what's one thing you're actively working to change? The best thing is working with a team that loves the creative industry and loves great creative work um to build tools for customers that are creatives trying to make great creative work. The problem that we're solving are like how do we help you make something that looks sick and that's really really fun.

>> So I I just really enjoy that. >> Uh the second part was like what would I change? >> Yep. >> I would want to have 10 more staff level engineers right now. >> [laughter] >> I would guess that's probably the answer any honest founder would give. >> Yeah. [snorts] Yeah. Very important. I mean engineers are the builders and without great builders, it's hard to build a great technology company. So that makes a lot of sense. Um switching gears a bit. What's one advantage of having such a diverse set of experiences early in your career?

>> Um I'm really good at metaphors. That helps a lot. I I I do think that because of a very diverse background um I'm able to make connections uh quite easily across different things I've seen like for instance for like our strategy I reference a lot um Chinese military history um because similarly to the AI space Chinese military history is different than western military history where it's really like stage battles you know let's fight you know all that stuff Chinese history of like a bunch of small folks popping up everywhere trying to take different segments highly competitive very very similar to the AI market today and there's actually you know surprisingly quite a few parallels like what land do I really want to make sure that we protect as the base and then go out from there and things like that so yeah and you know for product design like how might this feature work you know drawing the parallels of like the

personal computing paradigm and like basic and things like that to find like far-flung references that could help us today um I would say that's probably one of the main advantages >> yeah I love the reference to Chinese history I was actually an economics and Chinese double major in college so I studied quite a bit of both Chinese and Japanese history. So, at a high level, that resonates with me. I'll also say some of the smartest people I know, even in tech, studied history, and it's because history has a tendency to repeat itself. So, as long as you're creative enough and you can see the parallels, I think history is such an incredible blueprint for how to shape the future.

Um, next question. I'm I'm going to add lib on this one and go off script. So, you mentioned that you love poetry. What's your what's your favorite poem or one of your favorite poems and what do you love so much about it? It's a book of poems by Pablo Ruda called The Infinite Man. The reason why I like it so much is love was known for writing these really sappy love poems and he was really crapped at it too. But then he wrote this one book called The Infinite Man where it was very sparse um vaguely interstellar and very like abstract and nature oriented in a very awe inspired manner. not like super organic vibes, but like uh you know the cloud hung among the stars uh as the as the comets kind of rain down the crucifix of the mountains or something like that, you know, and there was no punctuation um no uppercasing, just purely just letters arranged. And the writing of it, it felt like there was a lot of space between

the letters, which I'm not sure how you achieve that. And uh it probably aligns most most with the type of aesthetic and poetry that um I would want to write and yeah, I think that's it's always stuck with me after reading that. Um so I would say that one. >> Yeah, I love that. And I don't know that I'm as deep into poetry as you are, but I I studied it quite a bit when I was in high school and college. a huge huge lover of Shakespeare and in in my mind the best writers and the best poets are able to do things with I was going to say the English language but really whatever their language of choice is that sort of completely reimagines in the mind of the reader what what's possible with language and the way you describe Nuda and how he creates space between the letters really just remind me of that.

>> Yeah. Yeah. Um, >> well, a couple more here before we wrap up. What besides Flora, what's one other AI tool that you use in everyday life and preferably one that Yeah. Okay. So, tell me about Granola and what you love so much about. >> Yeah. Where to start? I mean, one, they don't have they don't have a bunch of surface area and in some ways it's a highly, you know, people have done this before. >> Um, but I think like it show goes to show that like if you do something really really well, it'll stand out from all the noise. I mean, a couple innovations there that I think really set it apart. Um, you know, not having the joint calls with it. It's just on the background. It's recording. Um, I think the ability to see your notes versus the AI generated notes and, you know, the generations themselves being very good. That's all great. Uh, and the ability to easily canvas, um, you know, and ask your notes, things like that.

Um, I would say those things are what really make it stand apart. And, you know, all the small things, the way they handle folders, right? like uh the specific choices of what they expose and what they don't expose which is made more important. Um they have this new interesting thing called I think recipes which is similar to something that we're working on for like workflows. Um that's that hasn't been as good as I would like. Uh but yeah, I I think that's been it's cool to see that they're experimenting in that direction. Um, and they have like a recipe for like um, taking any notes and then like writing like a a PRD out of it and it's like made by Lenny. So, they kind of have like this name brand recipe. That's cool. I like that concept. Um, but yeah, I mean, you kind of took the words out of my mouth. Granola would have been my answer, too. I use it with most of my client calls and other meetings. And the

attention to detail and the craft that they put into it. Obviously, the obvious, the most obvious example being doesn't announce itself in the call, it's in the background, which is just so much more elegant and user friendly. Um, but also the the folders, the auto ads that I love the end of your waft they just did where they sort of >> give you the summary of your behaviors over the last year based off of your use of granola and it was so spot-on. It was a little bit eerie just how well they know me based off of, you know, all of the interactions they've observed in my meetings. So yeah, it's a great product and and I think a lot of the craft that they've shown in their product work um reminds me of how you describe your work with Flora. Uh last question is what's one way that being an artist has influenced your approach as a tech founder. It's influenced it for the better and for the worse. So at ITP it's

interesting because it's like using technology to make art. So like I would have four classes a week and every week there's like a assignment like you know make a make a film completely in a 3D environment you know using Unity that kind of touches on these concepts and you don't even know Unity at that point. So you spend the first half of the week learning Unity the second half of the week making a film for all four classes.

So what I learned from that was that um how to learn how to learn something new and then do something good with it. And I worked harder at, you know, my first semester at IGP than I did in investment banking. And the reason why I was motivated to do so and I was not getting paid investment banking money, that's for sure, was because I just really wanted to make something great. And that translated really well to being a a founder and like learning, okay, here's how you launch a product, here's how you build a product, here's the difference between design and engineering, here's how to like run a PM process, here's how to hire, like all that stuff. Um that was you know huge for that as well as the uh the focus on you know when you make art you have to make something people like. Um so I think that you know all the reps there helped us get really instinctually good at um making an interactive experience that people like

which you know we call product. Um I think the area in which it was a disadvantage was sometimes with art you want there's like a specific point in which you're like it's ready. Um product is like an artwork that is continually built upon for decades. Um there's no time in which it is ready. And the areas in which we've gotten caught up is when we we add scope or we want to make it exactly perfect instead of just getting it out there and testing the interaction, seeing it and iterating from there. And we've gotten a lot better at that over time. Yeah, I love that answer. And back to the Michelangelo analogy, you know, as an artist, you can sort of see the sculpture in the block of marble and human sensibilities or sense of aesthetics aren't changing rapidly. And so when it's done, if you're a great artist, you know it's done. But in technology, product market fit is a moving target. And with AI, it's a

faster moving target than it's ever been before. And so I think you're right. It's it's a piece if you're building a tech startup, it's a piece of art that's never done, and you just have to keep iterating on it. Well, this has been such an enjoyable conversation, as I knew it would be, Weber, and and in particular, I mean, I I love talking to all of my guests on the podcast, but it's always particularly wonderful talking to somebody who's so authentic to their mission, and it's very clear from your passion, your enthusiasm that you are one of those people. Any um final thoughts you want to leave the listeners with before we wrap up today?

>> Not in particular. Um if you're a creative, you know, try Flora. Let me know if you have feedback, I guess. >> Awesome. Yeah. Well, it's a great tool and I'm I'm just getting my hands dirty with it myself, but uh thank you for coming on today and telling us a bit more about Floor Story. >> Yeah, thank you for having me and for the great questions. >> Thanks again to RevenueCat for sponsoring this episode. If you're building a subscription app, you can use RevenueCat to power purchases, optimize pay walls, offer discounts, handle refunds, manage customer data, run AP tests, generate insights, and unlock more revenue. I work with dozens of subscription apps, and most of them are already using RevenueCat to power their businesses. Learn more at revenuecat.com/s subversive.