Transcript by AI

hello welcome to stack genius the podcast for gps and lps to create alpha with technology my name is sylvan i greet you from berlin today i'm also uh joined today by philip who are you in london or in austria where are you at right now uh right now i'm in austria but usually i spend my time in london Ah, fabulous. So, yeah, I mean, I think both of the companies or the things that you're doing are super exciting. So you have a fund sequel and I would love for you to tell us a little bit about that. But also today we will mostly talk about Pitchleak, a product that you have built. If I try to explain what it is, it's a pitch deck analyzer slash scoring AI that can help early stage startups or VCs to analyze cases. So how did I do? How would you describe what both SQL and Pitch League do? Well, you did pretty well. The one additional would make is that SQL is not necessarily a fund. We are a community of the world's best athletes investing in the world's best startups. And so the differentiation here is that our athletes have the ability to say yes or no and make the decision on each investment they make, which is a bit different from a fund. But yeah, I'm not here to talk about SQL. This is all about Pitch League. So let's jump right in. I think you put it very nicely. If you want, I can tell a bit about the intention of why we built Pitch League. Yeah. So as you already said, at SQL, we invest in startups. And I think as many investment professionals, we look at a lot of startups that come across our desks every day. And the question that we asked ourselves when ChatGPT came out especially was, well, how can we give some value to founders? Being quite new to the scene, we wanted to give something to founders so we make a good impression on them. And one of the very common tasks we saw and quite low hanging fruit is to give feedback on pitch decks, especially for early stage founders. Because sometimes a pitch deck just doesn't have all information investor needs. There might be some things missing. There might be some things you would not want to put in a pitch deck. And so we came across this problem a lot. And yeah, with ChatGPT, we saw this as super low hanging fruit to automate. Sounds super exciting. So how does it work conceptually? So you have some sort of interface. Is it a graphical user interface or is it an API? So how is it built? Actually, it's both, but for the founders, we have a graphical interface. It's a website called pitchlake.ai. Founders can go there, they upload their pitch decks, they wait a couple of, well, maybe 10, 20 seconds, and then they get presented the result and a full in-depth analysis of the pitch decks across many dimensions. That sounds cool. So, I mean, can we actually look at it together? That would be awesome. Sure. Cool. Then let's move over here. Here we go. Yes. So this is the website. And as you can see, it's quite simple. We've had already 3,000 and more founders using it. So we're very proud of that. You go here, you upload your PDF file, your pitch deck. And then it gets analyzed as a short waiting screen, but I'll show you the result. It's mostly suited for pre-seed and seed founders, just because the information that is needed and required in pitch decks in later stages can be a bit different. So really this is focused on pre-seed and seed. Here's an example, if you come to the website and want to take a look later, you can just click on it. I already did that. And so the feedback you get, we here have an iconic pitch deck of Airbnb. And I'll just go through and explain you what you can see in here. We have two major dimensions. The one is feedback about your deck. And then we have something very exciting as well for VCs, some scoring and ranking. So for the founders, probably the deck feedback is, and we get this feedback a lot, it's very actionable, very useful. Funnily enough, there's a couple of things we check in here, spelling and grammar. structure what it means is if all information is included if the deck is maybe too long or too short and then the clarity is just a score of is there too much information a slide or too little The best score is 100 or what's the scale? Here is the total score for the deck. And then those are sub scores. And then they are weighted with each other according to our formula. And then a final score is calculated. So Airbnb pitch deck, if you look at it and do a small review, actually wasn't that amazing. um the main reason is that it's just way too long right it's it's 26 slides which um with the attention span of of many investors today is a lot of information to go through especially in a deck in the first deck that you would send to investor if you maybe talk to them later obviously more information is always good but for pitch deck we we think the best amount of slides is around 12 to 13 slides Just to quickly check if I may. So I think you sort of said, you know, later stage decks function differently. So is this best use for teaser decks then not for full investment memos? Yeah, so again, right? If you're a founder in pre-seed or seed, and maybe it's the first deck you send to investor when you reach out, right? The goal of this deck, in our opinion, is to get a conversation started, to pique the interest, to include all the information that's necessary, but maybe keep some of the very interesting or very secret information also out of that deck. And for that reason, we designed our pitch league around this assumption. But maybe let's go through a couple of slides and then we can understand better how it works. So you can click here through each slide of the PDF so you can still see what it was. And then we give you actionable insights in what you can do to improve your deck and also improve your score. So, for example, this slide looks a bit boring, maybe. We could include some more data in here in the problem, right? So price is important. What does this mean? We know from Amazon, you should always use data, not adjectives. So as a founder, if I see this feedback, I could include some percentage, some statistic, right? How important for how many people is this important? Stuff like that. I actually do like, by the way, that it picks up on the missing data on the slide, which I concur with. It would be good to have data with a problem statement. And I like that it picked it up. Yeah, yeah. So we get this feedback a lot. Founders are always amazed at how well it works. And to be honest, we as well, because a lot of the magic is ChatGPT and our prompts behind that. So there's also content feedback, which just kind of tries to reiterate what's on the slide. And if there could be something added, again, it's dialing down here on the data and the statistics. But the score really would be influenced here by this. It also, and this is what we're very proud of, has design feedback. So it tells you, really, it takes the design of the deck into account. There are some investors that think that this well-designed deck is kind of a representation of the company and how they also treat design and branding in their company, right? Which for some companies is more important than for others. Yeah, for a dev tool, probably not. For a dev tool, probably not. It might be distracted even. Yeah, exactly. I mean, again, right, a dev tool. Also devs, I'm speaking here for myself being a dev. I like to use products that are beautiful, even if they are focused on devs, right? But yeah, it's maybe less important than for other products. They can be quite technical dev tools, right? And so it's not really maybe the first thing you would look at into scoring and stuff. Anyway, we gave some design feedback. And I think it's always a quite good summary here. You can maybe go through one or two more slides if you want, and then I'll go on to the scoring ranking. Is that okay? Yeah, please do. Which slide interests you most? Let's go for business model 12. 12. So, again, here, increase your score. elaborate on cost structure and revenue streams per transaction. So one thing that we really like to see in DAX is data, right? And so the investor really has all the information they need to make the first decision if they want to continue this conversation. Again, some feedback on the design and on the content. I don't think it makes too much sense going really too much into that. I agree. But maybe let's just continue with the scoring ranking. Do you agree? Yeah, I would love to see that. Perfect. So here we have the spelling and grammar. So if there would be a spelling, oh, there we have one, right? Spelling and grammar mistake. then it tells you this and also as well here. Sometimes this can be a bit nitpicky if you've got the feedback example here, right? It's all capitalized. Maybe that is a design choice by the founders. We also have the other scores. You're also able to re-upload your deck. So that's kind of one of the features we're all so proud of, right? That the idea is really that founders can use this as a tool to update the deck first, get feedback, and then once they made the improvements, ideally the score would be higher and this would be reflected in the score. Now let's continue to the next section. So here, again, we have the DAX score. This is what we saw on the previous slide. And we also have a score about the startup itself. We have here some information also that you can see on how those scores are calculated. We have attraction, a market, and a founding team score for the startup score. And this is how we roughly estimate the quality of the startup just based on the deck. Obviously, for Airbnb, we would have liked to score this higher, but maybe, again, not all the information was in the deck, right? So the structure here, that means that some information just is missing. This could be about the attraction, about the market size, things like that. And so maybe just not all information is in the deck, and that's why it kind of scores a bit lower on the startup side. And also... To co-comment, I mean, Airbnb did have problems raising funds in the beginning, and they did not see traction for quite a while. So I don't know when the deck is from, but it could fit that bill, basically. And also, I think it's going to be super hard to find the real outliers with an algorithm anyway, right? So it's going to be tough. We can get to this later. I can get into some statistics and I'll tell you a bit more about, we did some meta-analysis of all the DEX we received and looked at what information is included in DEX. So we can look at this later maybe. And I think there are some quite interesting insights in there as well. Cool. So you have also your ranking. It's called Pitch League for a reason. Right. So you compete kind of to get the best scores. We also extract the founders description and name and website and the website is available. Before we go deeper, just for my understanding and for the audience. So PitchLeak can also be used, you said, as an API. So for example, another VC could decide to send his deal throw through the API. Is that correct? Yes. So for VCs, we have two use cases. One is the leaderboard. So every startup that gets uploaded here, they appear on the leaderboard with the score. So for example, here, and then they can request intros to those startups. We just have to request intro as a filter for the founder, right? So they don't get spammed by just bots on the internet. And then the second like use case that you said, so yes, PeachLink AI is powered by an API, which we also use internally at SQL to analyze all the decks we receive and come across startups and we extract certain information. So for one, we compute all the scores, but there's also much more information that we can extract from the API. Just to give you some examples, we extract the ARR, we extract the MRR, the month-for-month growth rate, we extract the market sizes, we extract if the company is doing revenue, has revenue already, and a couple of other stats. For example, did they previously raise already a gross margin? previous founding experience, have the founders at an exit, and then also So you get structured data, basically. Yeah, like all structured data, like the region they're based in, the currency they use. So of course, we can only extract it if that exists in the deck. And that would be the segue into the statistics that we extracted about the startups. Yeah, please show. And while you open that up, so if they send their deal flow through to your API, do those companies show up in your ranking or is it private for the VC who uses the API? That's completely configurable. We do test it currently with a couple of VCs and by default, we don't show it on the leaderboard because like one of the things that VCs obviously want is they want proprietary deal flow, right? So if they're losing in a deal, they don't necessarily want to other people to know about it or they want to be very selective about it. And we also kind of segregate this data into its own database. So nobody at SQL has access to it, right? Because there could be some conflict of interest for sure. We at SQL also don't use the decks from Pitch League. We only look at the leaderboard and the scores and then we contact startups if they have particularly high scores. Just like any other user could request the intro. Exactly. Same thing, basically. Cool. Let's look at the statistics. We'd love to. Yes, I have them written down here. And so what we did is we looked at, of all the startups, so we received 11,000 pitch decks so far. And we have really startups that have used actually PitchLink more than 100 times already. Wow. Which is a bit crazy and not something we expected when we launched this tool, right? So what we did for those 11,000 startups, 11,000 decks, We looked how many unique startups there are, and the answer is it's roughly 3,500. And then what we looked, how many of those startups do include one or the other information in the tax, right? Because I said before we extract, for example, the annual recurring revenue, because there's a very interesting number usually for VCs to look at and to understand, well, OK, how far along is this startup? Are they growing? Which ARR? That typically gives you a rough estimate around their traction, around how well they're doing and how far along their journey they are, right? And so MRR and ARR together, so one of each, because you can compute one from the other, is represented in roughly 10% of all the DAX and Pitch League. So a lot of pre-seeds, pre-revenue companies probably. Well, I mean, there's two reasons for not including the ARR in the pitch deck for a startup, right? One is they don't want to share that number. And the other one, they don't have that. But we also have the number if... startups are already generating revenue. And so 50% of the startups on Peachtree actually generate revenue. So they would have AR, MRR numbers. They just don't decide not to put it in there. So if you calculate the 10 and the 50%, right? And it gives you that roughly 20% of startups that have revenue, put it in their deck and 80% don't. I find that quite interesting. It is, yeah. We also look at the founder's background and how many had an exit, for example, or previous startup experience. And so what we found in our data set is that 7% of the founders had a previous exit. At least that's what they say in the text. We didn't verify this data, but this is self-reported data. And 72% had previous startups experience, which is actually, I think, quite high and much higher than what I had expected. So that means they've worked at a startup or founded a startup before. That's interesting indeed. I mean, if we zoom out, thanks for sharing those data points, by the way. So if we zoom out a little bit, so... I really like what you built because it produces value for, you know, early stage founders, which can for you as a VC can be very positive for your brand, right? So if you provide value to them, they might remember you well afterwards, specifically if you're not Sequoia, or like, you know, whatever a big household name. And even I could imagine that you could use it for graceful declines in sort of lack of a better word. Even if you say no to somebody, you can give them something back. Exactly. Yeah. Very good point. So what we do is, for example, when we reject a startup because they either don't fit our investment focus or they are too early for us, right? We still want to give them some value because everybody in the SQL founding team has had a startup before and has gone through the painful process of fundraising. Decline, decline, decline, decline. Absolutely. It is a marathon, right? Yeah. So what we'd like to do is when we reject the startup, we send them some resources we've collected and Pitch League is one of them, right? To help them kind of on the fundraising journey. Cool. No, I like what you guys do. And I mean, just trying to take a lens on why you share this with other VCs. Is this so that you can possibly co-invest with them or why are you opening up this technology? Well, so what we like to do is, I'm a very big advocate, is if someone builds something once, why should other people have to build it themselves again, right? So that was one of the reasons. The other reason was just that VCs actively requested this from us, right? And so this is why we provided this possibility to them. and again right we for us it made a lot of sense to build this because again on the one hand we provide value to founders on the other hand we use it ourselves so every pitch that comes into sql is analyzed per pitch league and we can see all those stats in um in our tools and then we can also like build our filters or criteria for sorting and shifting for startups on that And then, yeah, the deferred use case with other VCs that just came up organically, they started asking us for that. So happy to help anyone or just connect and discuss, tell them more, answer any questions about Pitchlink or how about VLC fulfilled tools. Yeah. Cool. Now I would love both your product and your philosophy, uh, on that. So, um, yeah, well then maybe just, uh, in closing, the only relevant question to ask is what is good in Austria and what is good in London. So what are your favorite restaurants? Sorry, my favorite restaurants in London. Yeah. Oh, um, There is a really good Indian place. I mean, of course, Indian in London, right? Trishna is around Westminster. So I really like that one and I hope I can go there. I love Indian food for sure. And I love that an Austrian likes Indian food in London. We're all humans of the world, I always say, right? I guess so. Yeah, absolutely. Philip, thank you so much for spending time with us today. I wish you all the best and hope to connect anytime soon. Thank you for having me. All right. Bye now. Bye.

‍