The Demo Layer Is Becoming Core GTM Infrastructure

Machine-generated transcript from the episode video. Speaker labels are not included and some names and product terms may be transcribed phonetically.

[0:00] Most demo experiences are still screenshots in a slide deck or a 45minute call where the prospect watches someone else click. Joseph Lee built super demo super demo to replace both AI interactive demos used by 150,000 professionals across 100 countries. Forb's 30 under 30 at 22. First company scaled to 3 million in revenue. It wiped out 90% of it overnight. The second company went from 100K to 1 million ARR in a year, growing 3x a year over now growing 3x year over year profitably. In this episode, Joseph breaks down why the demo layer is becoming core GTM infrastructure, why most productled companies lose conversion at the moment they should be proving value, and what changes when AI audits your demo against 300,000 data points before a prospect ever sees it. Welcome to GTM Vault.

[0:56] Trusted by over 25,000 founders and operators building the future of revenue. Today's guest built two companies 30. The first was North America's first predictive marketplace for seafood. It raised 2.5 million and made Forbes 30 under 30. Then co destroyed 90% of the revenue overnight. The insight from the failure became the foundation for supera demo AI native interactive demos that help teams convert faster and onboard better G2's number five fastest growing product 2000 plus paying customer X growth in 2024 3x in 2025able Joseph went through tech stars he thinks deeply about why PMF has a short shelf life now why distribution and brand eminence are becoming the only durable moes and why The margin for shallow products is collapsing while the margin for compounding ones is expanding. Joseph, welcome to GTM Vault. Wow, thank you Rick. That was very well researched. Kudos to you. You built Fresh Line to 3 million in revenue and lost 90% of it overnight.

[2:03] What did that teach you about building systems versus building momentum? Yeah, I think one of the biggest learnings for me was being methodical about the type of problem you're solving and the type of market that you're going after. For me, I think as a naive young entrepreneur, I kind of dived right into the first problem space that I faced and uh just started building and and really being solutionsoriented rather than problems oriented. And and in hindsight, if I had looked back and had not been so naive, which is hard to do, I think uh when you're a young entrepreneur, I think I would have understood that there were infrastructural challenges within the market. And since the market in seafood was very traditional and hadn't changed in in decades, if not centuries, I guess we downplayed the amount of change management it would take to get people to adopt brand new software. And so looking back on it, I think the biggest learning was make sure you choose a market with the right, you know, headwinds where it's eminent, it's growing, and you have a real answer to kind of the why now uh question, which for us at Super Demo, it was very much

[3:08] the nexus of you know, new tools emerging and trying to um you know, demonstrate why they mattered and stand out when there's a new tool pretty much launching every single week. AI making it easier than ever to create, maintain and and share content. And then third being just PLG, productled growth exploding across the scene. Companies becoming more and more productled rather than salesled. So yeah, let the market kind of drag you along and help you build your company. You went from seafood logistics to interactive demos. That is not an obvious pivot. What was the structural insight from Fresh Line that became the foundation of Super Demo? Yeah, I think I I would go hearken back and go back to my initial point of like going into an emerging market and in a market where you can ride the wave. Um, and it makes your job as an entrepreneur building scaling much easier. And uh the critical insight for super demo was actually luckily derived from my time at Freshline when you know we had a very esoteric product very difficult to explain why distributors and wholesalers should start online ordering when it

[4:15] comes to taking orders from restaurants and different businesses and uh switch off of like facts, pen and paper uh and and just kind of analog traditional processes. So in that process, I had a lot of trouble trying to explain our product, why it mattered, the features, the workflows, and really the net benefit they would walk away with. But when I did kind of a more of a screen sharing back and forth interactive session just like this, that's when they kind of got that aha moment like, "Holy crap, I understand why this is valuable. How do I sign up?" So I had experimented with video, Loom, screenshots, different tools to try to kind of convey the same messaging, but it just it was so super passive. People didn't want to watch a 20-minute video was constantly out of date because our product was changing all the time. So um it was one of those problems that I knew, hey, creating demos, creating tutorials, creating onboarding documentation is essential for all kind of processes, but uh today it's just super super archaic and hard to do. So wanted to build a better way. You hit 1 million ARR in a year with Super Demo. What was the GTM motion at that stage and how much of it was things

[5:21] that do not scale? Almost everything was doing things that don't scale. I guess the first layer I tackled number one um SEO LLM visibility wasn't a big thing back then in 2023. So we went all in on SEO um across bottom of funnel, middle of funnel, top of funnel. Um it was all about let's create a large surface area for people to discover us, measure what is working across kind of the funnel and double double down on those things. So to give you an example on the bottom of funnel, what we ended up doing is hey let's let's latch onto existing demand for competitors or adjacent tools in our industry. People are already searching for Loom. People are already searching for top demo automation platforms. People are already looking for repris alternatives. So, how can we latch on to that demand by creating programmatic alternatives pages, listicles, and different pages that kind of insert our pretty fresh and brand new product into the category. So, that was number one. And then we would try to do that for as many different competitors and tools as possible and then measure, hey, which ones are actually picking up traction.

[6:29] Then, now let's actually craft it and make it more valuable and more detailed. So, that's what we did on the bottom of funnel. Middle middle middle of funnel was all about how do we actually show don't tell and and try to use our product which in itself is a tool to create tutorials and demos and workflows and try to create for example like product demos of other tools that people are searching for and again try to get ranked for a specific keyword there. So, if somebody's searching for HF's product demo or Canva product demo, how do we actually create an SEO optimized page that has our product embedded and gets people to the answer they're looking for, but in an interactive engaging way where that puts our product front and center and they can try it out themselves. On the do not uh things that don't scale side of the business, we employed that tactic strategically and as many places as possible. So that meant, you know, creating demos for people on Reddit, responding to product updates and change logs that I would get for popular product demos and responding with an interactive demo of the update that they could use, and just doing whatever I could do to make it easy for prospects to buy because I think a lot

[7:35] of software companies nowadays still don't don't realize that by making it as easy as possible, by making it white glove, by doing the work, by showing not telling, you're going to be able to stand out way better than someone that as you know book a demo form and forces people to go through a lot of their own manual work to adopt your software. You recently wrote that if your SAS only exists because it was hard to build once you were exposed. What makes Super Demo defensible in a world where anyone can ship software in a weekend? That's that's a great question. And I think there's multiple sides to defensibility and I think frankly speaking I think us including every other company even if they are defensible are paranoid to a certain extent and we're trying to invent and reinvent ourselves. So number one I think it's team I think is is the biggest factor. How malleable and how adaptable is the team to market shifts and shipping new products and iterating.

[8:30] But within the core product side I think it's all about um distribution. So for us, every super demo that is created, it organically brings in other customers because you create a super demo to share out with the world, with prospects, with customers, with colleagues, and once you have a critical mass of super demos, so for us, we're approaching 200,000 users, very much like Airbnb could, you know, replicate Airbnb today, but it's the breadth of people and the depth of people on the marketplace that create value and that stickiness. The same thing on Super Demo where we have folks using it laterally across multiple use cases not just in one category like sales but in sales, marketing, training, onboarding, enablement to the point where it's baked into folks kind of habitual workflows where to rip that out as you know you seldom like pick up a new tool and make it your daily driver every single week. It takes a lot of change management and sort of persuasion to switch off of tools. So for us it's all about how do we get in and expand laterally and also how do we continue creating this flywheel of 200,000 users where um you know we have our badge we

[9:39] have our premium plan we continue to bring more and more momentum and people uh into the fold and the last part is as you touched on earlier Rick uh brand eminence how do we how do we ensure that everyone in our category knows about super demo and we're we have eminent domain across demo automation so somebody thinks of interactive product demos Super Demo is one of one, two, three names that come to mind and very seldom are people looking beyond the top three or top four or top five. I want to dig deeper into this idea of how most companies treat demos as a sales tool whereas you're positioning Super Demo as more of an infrastructure. Now tell me exactly what is the difference and why does it matter architecturally? Yeah. So, what we try to really encourage folks to do is look at Super Demo as a living and breathing asset. I think the mistake a lot of people make is they put a lot of upfront work into creating a demo and and using it as a perishable asset, right? But the way we've methodically built out the tool is everything is modular and everything is like composable and reusable. So that same demo that you've created can actually be slightly altered, slightly modified or used asis

[10:52] across all the different areas. So essentially maximizing the surface area of where customers are interacting and learning about your product. So instead of just using it as a sales demo leave behind, you can use super demo by embedding it in your email sequence. You can put it inside of your app as like interactive product tours for in context enablement. You put it in your support docs, download it as a video and use it in trade shows or obviously use it in marketing, product updates, change logs, etc., etc. So, by making it maintainable and reusable and turning Super Demo's kind of um content library into the source of truth, you're now giving one place where people can make updates, tweaks, and and and create offshoots of content, changing essentially all the different areas where that demo is being used and embedded. So that's the first step. But the next step beyond that for us is that we're building out here in Q2 is how do we now use agents to automatically create automatically maintain automatically redo demos that are out of date. So that that mental anguish of you having to remember oh this demo went out of date or this demo

[12:01] is used in this place or that place. Uh that's abstracted out and handled by agents on your behalf so that people have the right up-to-ate content 100% of the time. And you have data from how many demos now? Oo, I don't even know off the top of my head. Probably like 3 400,000 demos uh across the board. Yeah. And what patterns separate the demos that convert from the ones that don't? Yeah. And that's actually another common fallacy. I think a lot of people the mistake that they make is they spend a lot of time building one hero demo that they try to squeeze in everywhere. And Rick, as you know, at every stage, even in the salesunnel, you need to have different messaging and you need to have different like breadth and depth of content. You can't just try to squeeze every feature and every benefit into like a top offunnel marketing demo. And nor should you have like highle content in a very bottom offunnel like conversion focused content. So instead of trying to create for everyone, the biggest takeaway is make sure the length, make sure whether you use voiceovers or text or even like the language of whether you're benefits focused or instructional actually maps to the use case that you're optimizing

[13:12] for whether it be customer success, whether it be top ofunnel marketing or bottom of funnel sales. Right. And not to mention the stakeholder who it's for as well, right? Yes. Exactly. You launched AI demo audit which scores demos against best practices automatically. Walk us through what that system actually evaluates and what it catches that humans miss. Yeah. Yeah. So I think um the the key insight here was that we found that we had a lot of really good resources and data around what makes a great demo for for a particular use case and a particular funnel stage. But all that information was scattered throughout our documentation just internal tribal knowledge that I or you know our head of customer success Paulina would have. Um and it would just be on our academy. So it would be up to our customers to go out find that data watch that information or watch that video or or uh absorb that information and turn that into insights that they could actually apply. Uh what we ended up doing is hey how would it um how would customers find value if we could connect all that kind

[14:20] of silo data put it into one trained model um on best practice demos the top performing demos our video and our written assets and based on the use case that they chose whether it's marketing or um sales or training and also the intent that they want whether they want more completion of the demo or they want more reach for the demo and the context that they prov provide for that demo itself. If we could audit and give them recommendations on exactly what features to use, how to change the hotspots, what to add to better align with their goal, but not only that, not only giving them the score, if we could give them that one-click ability to apply those changes so that they don't have to go in manually change things. That was the intent. How do we get them to value and how do we get them to the destination faster without them having to do all that manual work up front? Interactive demos sit at the intersection of marketing, sales, and onboarding. Who should own the demo layer inside an organization and what breaks when nobody does? That's a great question. I don't think I have like a perfect answer for you there. I think a lot of that depends on capacity and who is the champion and who

[15:28] feels strongly um about owning it. I think I've also seen I've seen people in customer success own it across their teams and spread it across the revenue function and the sales function. I've had sales directors own it. I think that depends organization to organization. What I will give you is Rick that yes when there is a little bit of dysfunction and unclear hierarchy of who owns what, what the process is to create a demo and what the process is to get share and enable and embed existing content. Folks typically don't get as much value out of the tool as they as they can. So yeah, that's that's a good question. I don't know if I have a a right answer for you, but that's probably one of the things that we're frankly working on is how do we better enable our champions and how do we select the right champions so that our tool gets adopted because that's that's the biggest hurdle, right? Change management, getting people to adopt creating is easy nowadays, right? The the latter is is a hard Yeah, absolutely. I mean I I presume based on um company maturity uh and and stage of of um of growth the organizational structure is going to be

[16:39] different. Hierarchy is going to change over time. It's a hard one going to change based on a lot of a lot of factors. You grew 8x in 2024 and 3x in 2025 profitably. Most companies at that growth rate are burning cash. What is the structural decision that made profitable growth possible? Yeah, I I think uh capital efficiency has been at the heart of everything we do from the get- go. And I think from myself and my co-founder building the company initially for the first say 8 months, just the two of us all the way down to us being you know a 16 17 person team today. Um I think that capital efficiency and and that frugality in a positive way I think it's been instilled as like a cultural tenant within the So we try to do everything efficiently. We try to keep profits and customer value at the forefront of everything we do rather than posturing or thinking about how we're going to be perceived with investors that we we see as like a um a tertiary or side benefit to creating value for our customers and being building a profitable capital efficient

[17:48] enduring business. For me, I think the reason we've had that mindset is uh you know, one of the mistakes that I made with my first company was raising too much too early and raising too much when we didn't have the right PMF levers for long-term sustainable growth. So, what happens when you do that is you end up chasing the next round and indexing and optimizing for the investor narrative and the path towards the next round rather than the path for sustenance. So you end up raising too much money. You run out of optionality for you and your employees because now you have this like huge preference stack you got to clear. You can only sell the company for a certain amount of money. Um or you need to continue down the venture path. So I think it it I realized the next time around not to say you shouldn't raise money or you shouldn't know take one path through the other. I just wanted to create an environment where we controlled our destiny and the only way to do that is we build a profitable capital capital efficient and uh enduring business. You said PMF has a short shelf life now. You ran your own version of the cursor

[18:57] all hands and asked your team what assumptions might already be outdated. What came out of that conversation? Yeah. Well, I think um if you doom scroll tech Twitter, I don't know if you do, Rick. Uh I think you always feel like there's an existential flaw or or a finite timeline that you got to figure things out as a startup. But yeah, I think when we did it, it was all about right society is shifting as a whole. Things will take a little bit of time. We also sell to a portion of our customers are in more traditional industries where it's going to take time for ch change management. But I would rather make structural changes if needed or reinvent our business if needed in a position of strength rather than in a place where we need to pivot or we need to change or we need to morph. So the biggest takeaway for us is cursor for example is hitting $und00 million in ARR plus. They're growing exponentially and yet they see structural fundamental existential threats for example from cloud code where people may be not dealing with an IDE or a code editor in the future. They might just be prompting everything. So they're thinking about that space and figuring out how do we

[20:05] ensure the long-term durability of the business even when things are growing at breakneck break neck speeds and and and um going so well. And so I wanted to emulate the same thing where hey we are growing really quickly. We are very profitable. The product is being adopted across the board. But how do we think about the business and the space in five years from now and how do we structurally put ourselves in a position where we are nimble enough to make changes as things change. No one can predict the future. Right? So the team kind of took away the biggest message was, you know, forget about everything that made us successful in the first place and be malleable and open to just ripping up and tearing up the the playbook that we built and being open to reinventing our growth, our product, the way we've done business because we may have to. Not to say we will have to, but we may have to. And uh people need to understand that mindset. You are at 9% month overmonth MR growth with 3% churn. What is the biggest structural constraint between where you are now in the next phase of scale?

[21:07] Yeah. Um I mean as we march onwards to you know five and 10 million in ARR I think um as funny as it is it's it's shedding some elements of what has made us successful as a startup of everyone doing everything and and there being just complete structured chaos I would say that we've uh dealt with and having a little bit more process so that the entire machine can move in the right direction. I think that is one key and I think the other thing is just doubling down on what we have done well. So the biggest focus for us is you know as we transition as a product how do we become more agentic. So we're building a ton around you know AI demo agents MCP AI editor uh there's a ton that we're doing there but in addition to like what we're doing within the product how do we continue to build with change management in mind because again that's actually going to be an even bigger problem than before because there's more tools and more more things vying for people's attention. How do we get embedded into core workflows and the way that people do business in a seamless way so that you know super demo is anywhere and everywhere in someone's day and it's

[22:18] impossible to pull out. That's that's the core problem that we're looking to solve moving forward. You wrote that distribution and brand eminence are becoming the only durable modes. You started a YouTube channel from zero. You post consistently on LinkedIn. How do you think about content as a GTM system, not as a marketing channel? Yeah. Well, I think it's essential, hearkening back on what I said, I think there's more noise than ever, and it's harder to earn people's trust. And the only way you earn trust is storytelling. So, the right storytelling and consistency. You can't just post once on LinkedIn and expect people to trust you and believe in your product. You need to be doing that day in and day out, right? So, I post on LinkedIn probably three or four times a week. And I've done that for like two years now. And um I've gone from like 500 followers to almost 20,000 followers as a result of that. But more so than the followers, it's built a level of familiarity with Super Demo and a level of trust where, you know, I'll go to a networking event or a founder event here in New York and probably like three out of 10 times folks are like, "Oh, I've heard of Super Demo." Or, "Oh, I've seen your stuff." Or, "Oh, I've heard of it through the grapevine." And

[23:24] I don't think that happens without you being a megaphone across as many channels as possible and just like yelling and telling your story from the rooftops as much as you can. And so I think in order for you to build a durable business and again be one of the standout products or companies within your space, you need to be posting anywhere and everywhere. And not every channel is going to be effective for everyone. Somebody might benefit more from Tik Tok versus YouTube versus LinkedIn. You just got to find your space and and tell your story consistently. You use AI internally to prototype and amplify product building. Where does AI genuinely accelerate your development and where there's a breakdown past the demo phase? Yeah, I mean we use AI um across the board, right? And that includes more one-off like use cases like cloud code, cursor to build things, but we also use agents that are running continuously that help us across customer success, closing tickets, extending trials or even creating dashboards so that we can predictively help score customers and create actions for people that are hitting certain milestones or certain triggers within our database. So we

[24:33] encourage and and somewhat mandate the usage of AI across the entire company because it'd be naive for us to be an agentic and be an AI native platform and not build with it and not encourage the use of it. So yeah, we we use it across the board in all all cases. Um I would say where it fails, Rick, is if we treat it like a silver bullet, right? A lot of people treat it like a silver bullet where you give it one prompt and expect it to produce the output you want. It takes a lot of craftsmanship and a lot of iterations and back and forth to translate what's in your head here into machine readable prompts and then from there even going back and forth creating guard rails and processes and refinements to get to the end state. So don't expect that it's this like magic wand where you wave it once and you get the output you want. you need to work with it like a real human and and craft over time.

[25:28] I want to dig a little deeper into change management and how the demo category is actually shifting actually. So we we talked about how change management is the most underestimated factor in AI adoption. If that is true, what does it mean for how fast the demo category is shifting? Yes. Um I think um there's always going to be a lag. Technology is always going to be uh a few steps ahead of uh human adoption. We just can't change and mold and adapt at the pace that new tools are coming out. Right? So to give you like a tacid example of that is Microsoft Excel, right? A triedand-rue product. It's been the way it is for a very long time and people have shaped entire workflows and habits around that tool.

[26:19] There have probably been four, five, six, seven, 10 different versions or iterations of Microsoft Excel from different companies that do get to the outcome faster or is a more powerful tool. But because of the change management and the the habits built around that ecosystem, it's been still very much the dominant player in the space. And now you can imagine when somebody does move on to the next version or the next product that claims to be Excel and they start using it, they're not going to just shift off of that after two months after they've gone through the training, they've gone through the procurement within the enterprise to adopt the tool, they've paid for the tool, now hundreds, if not thousands of people within a big organization started to use it and started to create output from it. they're not going to just be able to like change to a brand new tool two weeks or two months or even a year from then and restart that entire process. So people forget about the invisible cost of adopting software when it comes to the procurement the training the infosc the the change management and the actual adoption within a company that is a very very very high burden and that's why again I I touch on getting into habits getting into key workflows and getting

[27:31] into not just one person but across the entire organization across different functions because if you become the way they do business there's a cost to them changing and that cost is exponentially higher if they're changing often. You said taste is the new moat that the judgment to know what deserves to exist and what to leave out matters more than shipping speed. How do you build taste into a product team? That's very hard honestly. That comes with tribal knowledge of understanding your space inside and out. It very much is a skill curation skill that takes a lot of time and a lot of experience within the industry. It also depends a lot on what you're absorbing because we are a byproduct of like what we see, what we hear, and what we absorb. So someone that has good taste is, you know, ideally very very has a good good kind of ear to the ground and has a good pulse of all the good trends and all the new products and all the things that are happening. And how do you then take all of those inputs, you know, run it through a seieve or run it through a filter and pull out the necessary parts that are actually consequential to your business and implement them. It's no

[28:39] different than, you know, an excellent product manager back in the day where they would be, you know, in tune with all the trends and understand where the market is going and collect as much information as possible, but they're able to distill that down to actually what matters. And the same thing is happening now just at a much greater scale. Yeah. Uh the people that were excellent PMs back in the day I would say are are actually excellent taste makers. Uh today if you were starting Super Demo from scratch today, what would you build differently? A lot of different things. Um I would obviously build agent first, right? I think that's where the future is going. I feel like excellent products are going to have to have two interfaces. one for agents and one for humans that maybe use agents or use AI to get to the outcome they want, but visually they want an interface that is a little bit more deterministic that has a few more guard rails of what you can and what you can't do because when you're prompting AI, you can't control the fidelity of exactly what you want. Sometimes you get a slightly different output every single time, but in reality, humans want the same output, the same outcomes every single time. So, I would build for agents in mind from the get- go. How do we create a a layer in an ecosystem

[29:50] where we make it as easy as prompting on cloud code or on your machine to create a demo of a new feature or when a linear ticket gets closed automatically creating a product demo that you can then use and tweak for internal or external use cases. So, just making that entire process interconnected and available for any agent or any tool to be able to leverage. like to move on to the rapid fire section of the pod. Once first instinct, SAS is dead. True or false? False. The most overrated growth metric for early stage founders. Posting on Twitter. The one thing most product demos get wrong. Building for everyone instead of for their specific audience. The belief about AI in product that is completely wrong. That it is going to fully displace all forms of SAS across the board. The hardest lesson from losing 90% of your revenue overnight. Don't associate your your happiness with the company's trajectory. Vibecoded product versus compound product. What wins?

[30:52] Compound product every single time. You've built two companies from zero to seven figures. You've watched one collapse and built the next one to be structurally durable. If there is one thing founders get wrong about the relationship between product and distribution, what is it? Distribution is king every single time. That's the first thing you should optimize for. Product should come second. Now, more than 150,000 professionals, 2,000 paying companies, 100 countries profitable. The demo is not a sales artifact. It is a conversation system. The margin for shallow products is collapsing. The margin for compounding ones is expanding. If you're building GTM under real constraints, explore GTM vault. Build architecture not activity.