SaaStr AI Live: A Hands-On Guide to SaaStr's New AI Tools with Jason Lemkin

Next year this will be like 80% AIs and 20% humans. But welcome. Welcome to both our, our AI and humans attending from all across the globe. What I want to do today is we have built. Some of you have probably seen it, some of you have played with it, others haven't. A lot of tools on the new Saster AI we have built particularly a lot of tools around helping for fundraising and start evaluations, which I mainly want to show you. I think a lot of you have used our digital Jason or AI mentor. I can quickly show you that if you haven't, it's been used over a hundred thousand times. You can ask it any of your questions. And we've gotten a lot of folks using just our new pages on AI agents. A lot of folks ask, what AI SDRs do we use? What AI BDRs you? What is our stack of these 12 AI agents that Saster uses to run Saster? They're all here. If you click on AI agents on Saster AI so we can touch on all these so you understand what we do. But mostly I want to show you our tools and to the extent we have time at the end, all of these we've built. Vibe coding, it has taken a lot of time, but if folks want to ask questions behind it, how we built it, how it works, what we did, all the time we have at the end, we can talk about that or these tools or anything else. I'll try not to take up too much time on the tools, but enough to help you guys. And listen, the first one we built, I want to talk about our AI VC tools, but the first one we built is our digital JSON. If you click on AI Mentor on Saster AI, you'll see it will come up. This has been trained on all of our Saster content since inception. And you can ask any issues I haven't. And many folks spend almost their entire day on this. So, for example, I'll talk for an issue of the week. Let's see if I can get a CFO. Our CFO has missed cash projections 4 times this year. Will end the year with 4 million less cash, cash than projected. I actually don't know what the answer is. Zoom. What do we do? Okay. And essentially this will use our AI and it's basically incorporated every single speaker, every single section we've ever had. And what does he say? 4 misses on cash projections and 4 million less cash to plan. That's a huge red flag. You need to get crystal clear on why it's so bad. Is it poor forecasting discipline? Lack of real time visibility? You need to build a rigorous burn plan and burn rate budget. Review it weekly with the team. Everyone needs to know Runway. Get a rolling four month forecast that's updated weekly or bi weekly. It can't be so optimistic. You need a plan to either cut or mercury, raise more capital or both. How much Runway do you have left? Should. Okay, another question. Should we keep the cf? Should be zoom. Should we keep the cfo cfp? Let's see if I can handle that error. Four misses. It's serious problem. Sometimes what you need is a strong VP of finance or controller. You can keep on going on that, right? See, we can try another one and then we'll move on. So I'll try another one. We'll hit 20 million AR this year. 80% from self serve and 20 from sales led. But. But only one person. Zoom. But only one person person on our sales team can close. And our VP of sales never worked out. What do we do? Let's see what digital Jason says. This is a very anonymized version of a startup I know. And only one closer. It's a classic inflection point. You're scaling fast but your sales team isn't scaling with you. And your VP of sales didn't work out. You need a strong battle tested VP of sales potentially. Don't settle for one you don't love or trust. Start building out the sales team org structure now. Lean into your self serve strength and you can input resumes for folks. You can share their background. See if there's a good fit for you. So anyhow, this stuff's pretty good. And it's, it's. Frankly if nothing else, it can be cathartic for you. So this was the first agent we built. It's cool, it's free. You can use it on Sasser AI. You can click here at the bottom. It's also on sasser.com it's on all of our properties. So that's where we started. It's pretty cool. And then we built this aivc and, and I want to show you these tools because even if you're not fundraising or anything there, there can be a lot of value for anyone. And so just, just click on a VC and you'll come here. It's always at the top, even on mobile. And it's, it's done a lot. And let me walk you through it. It now can connect you with over 400 top B2B and I VCS. We're actually adding more Amelia is reaching out to several hundred that are coming to Europa AI London, which is better. You better come. It's December 1st and 2nd in London. We're going to do all agents all day, but we're going to add hundreds more this week. And they're here, folks that you know. Altos Ventures, scale Ventures with N20VC, with Rory Me on our pod 0.9, Bloomberg beta, which did Replit and many others. They're all here and you can connect with them via AI. It's pretty cool. And I'm going to show you how that works. And there's basically four core tools started with one, and they're all kind of related to each other, as I'll show you. They all build on similar elements and core, but they're different insertion points depending on what you want to do around startups. And let me walk you through them and then I'll show them all to you. The first one we built is. Is this just seemingly simple startup valuation calculator? What's your startup worth? It's a question that lots of founders and executives have, and I'll show you that one in a second. That was the first one we launched a couple of months ago. As you can see, we've done 569,000 calculations on it. Now, that's not valuations. Founders and executives will do more than one valuation per session, as I'll show you, but it's probably 50 or 60,000 folks have used this valuation calculator. It's fun. It shows you what you're probably worth, at least optimistically. And anyone can use that. You don't need to be fundraising. You don't even need to be the CEO or a founder. You can just try to figure out what your startup is worth. So everyone can have some fun with that. It's free. The next thing we launch, which I'll show you, is a pitch deck analyzer. If you. If you think you might want to raise from VCs anytime, you can upload your pitch deck. You can do it anonymously if you want. You just update it, we'll score it for you, which I'll show you in a minute. We'll tell you the odds you can get funded and we'll even potentially introduce you to VCs if it's the right fit. So you got a pitch deck. Don't go in blind, don't just send it around, don't email it to random people first. Come here, upload it. We'll give you feedback. You might not like the feedback, but we'll tell you what to fix and you can do it again and again and again and keep getting your pitch better. The third one I'm going to show you is a variant. Is what if you're not fundraising but you want to know how, how well am I honestly doing? Right. No one's really honest with folks, everyone. Most founders think they're doing better than they think. Some. I used to think I was doing worse than I thought. Now you can just upload an investor update, a pitch, a board pack, any of those things. That's what's different here. Anything. And we'll show you how you're doing. I'll show you that in a minute. You can research VCs and then the latest thing we've added and you can see it's kind of a combination of these things. When you're ready to race, when you're ready to fundraise now we will send us your deck. Our AI will see if it's honestly good enough. And it's a tough grader. Realize it's tough. We've talked about this so much on Saster, on the blog, on our events, on our podcast with Harry and Rory. I mean more we are. Vcs are investing more than ever, but in fewer and fewer higher and higher growth companies. It's tough. So we're going to give you an honest grade you might like, not like. But if you do like it, you can check again the VCs you want to meet with. With this new VC batchmaking. This, this one we just launched, but we've already made 168 BC intros, so try it. It's pretty cool. So, so let me walk you through these. The startup valuation calculator. This is where we started with the 570 valuations. And I'll just refresh this one. We'll go straight. So you come here. It's super easy. You don't even have to log in or do anything. Let me use a few examples from my portfolio. So I've got one investment I did. They're at series B. They're sort of AI enhanced SaaS. They're actually at about 25, 26 million ERA now they were, they, they're accelerating. So at the end of this year they're going to be at just about 100 million and click valuation. Okay. If they're really an AI enhanced SaaS, if they're in a hot space, they could be worth as much as 383 million or more. That, that's pretty accurate. But earlier in the year they were at about 20 and you can see they'll dynamically change it which is pretty fun. They're at about 20 and growth had slowed to about 70%. Still look still really good by most standards. But again the bar has gone up. Adventure. So what would they have been worth? Only 166 million which isn't much off the last round fundable. But it's going to be hard. And so not only do we tell you what you worth, not only can you download a whole report and share it around, we'll tell you multiple but we'll also tell you what the odds you're going to get funded based on the all the top 5,000 deals and all the top benchmarks here. So 70% still at a decent chance. As you get 100% the funding rate goes up and you know, boy, you push that up into the 150 and higher then pretty high chance you're going to get funded. So, so, so that's one I could think of. Let's try another one to seed stage. Coming up on a million in revenue now the bar has gone up now if you're, if you're growing 100%, a lot of founders think that's pretty good doublings. Doubling is great but you know, even triple, triple, double, double isn't good enough for most VC firms. Unfortunately. It's the reality. We've talked about it. You might be worth 20 million, you might be God, I'm only worth 20 million. But that's 20 times revenue. Okay guys, but what happens if you're growing 200% which is really 3x okay, it's pretty tough to do now I'm in that 50 to 60 million and what if I'm growing insane level? Well you just might have that insane valuation. So it's pretty good. It tells you roughly what you're worth, it tells you why, it tells you the odds you'll get funded and it'll show you all the benchmarks. Okay, If I'm at 1 million and I want to be funded and I know this sounds tough, what is a top quartile funded company not look like not all startups on planet Earth, but the ones that have raised venture funding this year, the Top quartile is 481% growth. It's pretty tough. We'll tell you that, we'll tell you where you come and that's kind of your odds of getting funded. We'll explain to you how it all works if you want to see it and then we'll show you next steps. You can do that benchmarking, which we talked about, or you can just get your deck reviewed. This is free. It's fast. You can do it every day. You can let your team do it. And again, people have done it 470,000 times. So that's where we started this. Use it. It's fun. Go to Saster. I click a VC and have fun. And we're continually updating this with the latest data. It's not perfect. In fact, we just pushed some updates this morning that'll go out with more and more data. It may be too optimistic if your growth isn't that great. It definitely reflects Bay area multiples, right, and the AI frenzy. But that's who's getting funded. It doesn't reflect who isn't getting funded. Like, we don't have a lot of valuation data on deals that never happened. So there is a null set, which I am thinking a little bit about. How do you value something that can't get funded or can't get sold? Well, does it have any value if it. If you can't sell it or fund it? Probably not, but, but. So in a sense, these valuations are all inflated because they represent the ones that are getting funded. But on the other hand, it represents 5,000 rounds. 5,000 venture rounds. That's pretty cool. Okay, so that's how to figure out what you're worth. And let's go back up here. The next thing you might do is get your pitch deck analyzed. And what you'll do is it'll go through all of it. It'll tell you your traction score, your quality score, and how good good a grade you are. And really, all you ultimately have to do is upload a pitch deck. So let's take one and let's see. Let's try this one. Human AI. Okay. Not quite a real company, but we'll upload it. It'll take a minute to go through Claude and extract all the data. If you want, you can give us your ARR, your growth rate, your funding stage, your company, type your name, email. It's great. But if you don't want to share anything on your pitch deck, it's okay. AI will do the best job it can. Right. It might not be perfect, but it will attempt to extract all of the data that it can and figure out how good your pitch is. And I'll show you some data in a minute. It's. I think it's pretty good. I've run all of the startups I work with that are Raising money or not raising money. I've run it throughout. They're pretty directionally correct over many years. Some folks don't like this. It'll come in a second. It can be tough. We used to give a lot of F grades and that was a little triggering for folks. So I think we changed to nf not fundable or nr. But honestly if you're not going to get funded it's just going to tell you because there's few things that are I think a worse waste of founders time than trying to go off and get funded and try to meet with a bunch of VCs if there's just no chance, a small chance, it's definitely worth it. You know there's so many stories of I met 100 VCs and I only needed one. So here's. So here's human AI VC. Now look and this one's pretty good. Okay now this is a synthetic company. I made it up but it's growing 900% at 1920 KR. So coming up on a million that's pretty good. That's top quartile. Not that the Top quartile is 521% so it's well above it. You can actually click here and see what all the top quartiles are across all startup stages Based on Iconics data, one of the top growth funds, all of all the venture funded companies they track and B2B here's how quickly they are growing at these stages so it judges that it'll tell you your traction score. You can click in and see why how good your deck is. This is pretty good. So for these guys they're track they've obviously they're crushing on traction. The deck's only pretty good. I made it up and then stuck it into gamma and made a really pretty deck out of it. But you know maybe if it's real we would have made it even better and it'll go even into more detail if you like. So down here it'll tell you how disruptive the company is. Is this really changing the world with AI? Because this is what VCs are looking for. I mean pretty good, not great. It actually you will actually give you a little bit of hit for not being disruptive enough. We'll break down everything from your deck quality to your team size. Up at the top we show you how good you are. But if you want to see how how good you are across all different types of benchmarking, we'll show you how good you are. Across Carta, across Bessemer, Benchmarking, across our own, all the, the analysis we've run through this and across the, the iconic benchmarking. It's all here. And here's the cool part. And we'll, we do we reuse this element. If your grade is high enough, we'll automatically ask if you want to meet with VC funds right here in the pitch technology. You don't have to do it. You can come back later. There's no pressure, but if you want, you literally can come in. There may be some you've heard of, like Saster Fund. You can click on it, you can learn about who is the fund, how much do they fund, who are their top companies, what's their philosophy? You, you know, here's Harry Stebbings and 20 VC. Here's here's others and you can click into them, you can learn more about them and if you're ready, you can click and we will go out and introduce you to them. We'll go out and introduce you them. And right now we've got about 20 top funds and emerging funds in there. But soon there'll be hundreds of folks and we as I'll show you later, maybe in the next time we do this, it'll dynamically pick the best ones for you. But it's pretty cool. It works. We've done maybe 500 intros through this and 150 directly through VC Matchmaker. So this will automatically come up. You don't have to do it. But if we think you are fundable, we will do this and show you some funds where we think you're a good fit. Are you good fit for Rory at scale? Well, is A and B $900 million fund, HubSpot build, DocuSign box. I mean, pretty good. It's all here and you can play with it. And there's nothing you need to do. Keep going down. The richest one we do is this pitch deck. We'll show you things you need to fix. And if you log in here, we'll show you exactly how VCs see you. And we'll even show you how VCs will write up their investment memo, which is pretty cool. So that's how it all works. Here's how all the benchmarking works. And down here we'll show you how you fit in. This one's a B plus. B plus is good enough to get funded again. Remember, we're a tough grader, guys. Don't take it personally. Here, here's the distribution. You can see there's a lot of apps, which just means they're not fundable. A lot of Ds, but B plus is pretty good and we've done 1822 decks so far. So this is fun. You will learn a lot. You can get a lot of feedback on your deck. I really don't think there's a better everyone. You know, you should always share your decks with other founders and other executives before you go fundraising. I can't think of a service that is better than this where you can do it with AI and online. It's, it's pretty cool. So that's the pitch deck one. The next one I can show you. Let's try. This one is benchmarking and it's, it's. They're similar but different. So let's see if we can try. I think I did Human test company name Jim. Jim at Human. He's a good guy. Jim Larkin. Jim. Now let's see. I think I've got one for Human. Here's their investor update. So what's neat about benchmarking is you don't have to be, you don't have to have a pitch deck. It's okay. You can. This is literally an investor update if that you send out every month. If you just have a board pack like people usually make board packs. You don't have to edit it, you don't have to update, you don't have to turn to adventure Pick. You can just take your board pack when you do board meeting and throw it in here or take an investor update when you've sent it out and get honest feedback on how am I really doing. And it'll give you pretty similar results to the pitch deck and benchmarking, but it has a few other cool features there. And again, the point of benchmarking is benchmark yourself every week if you want all the time. No one does this. And I find no matter what founders I talk to, they're all wrong. Like they're all wrong. They think they're doing better than they are. They think their cash will last longer. They think there's fewer issues. So I uploaded this investor update. They're growing 251% at a million in ARR. They'll show you how good that is relative to other folks, how fundable they are, recommendations on what to do. And this is pretty cool down here. If take your deck, it'll predict out your revenue and your cash burn. Founders do this all wrong. So right they'll it'll take your trailing growth rates at 1.1 and it'll predict out where it thinks your revenue will be one year from now. 3.6 million. Pretty cool. I know you could do this yourself, but people do this wrong all the time. It will also take your burn, which is at 200 minus 200k a month for this startup, and it will predict how that burn will grow and where your cash position will be. So. So a year out, it's saying you're only going to have 600k left in the bank. Maybe we want to do a little fundraising, but everyone gets this stuff wrong and we're actually going to do even more here. But it's pretty cool that it'll project out your revenue and your burn rate just so you can see. And again, it'll also bring in this benchmarking for this startup. How am I doing compared to all the Carta startups, Everyone that Bessemer tracks, the shooting stars, everyone that sat, everyone that's ever uploaded something to Saster, which is probably the largest data set now. And again, this iconic for the top quartile. So, you know, being this isn't being in 52% of the top coral. It's good, it's very good, but it's not off the charts as you can see for all, all the bars down here below. So that, that's benchmarking. It's a version of the same thing. It's pretty cool. And then I'll just show you the fourth version. If you upload your deck to VC matchmaking. Let me show you this one. Hold on. They're all pretty cool and you can see they're similar but different. Here's the newest one. Find the right VCs for your startup. It'll basically take you through the same flow, but a little bit different. We get a little bit more data. When are you fundraising? So look, if you don't want the pressure, if you just want to take coffee meetings and get to know VCs, it's okay. If you're phone ready this second and you want meetings this week, click that. If you just want to learn, that's okay. And even if, like you don't want to talk to any VCs today, but maybe later you can be kept in the loop. You fill that in, what stage you're at, whether you want to meet with angels or just vcs, how much you're raising and just upload your deck and you just go from there. It's really, it's really quite similar. And. Okay, hold on. Data test. Get good old, good old Jim back and then we'll just upload it and it does it for you.

Hey Jason, while you're doing this, there's a question in the chat. If you retain the data from the pitch decks that have been uploaded.

We only retain it if you want to retain it. It's a really good question. And if you get a high enough score that you want funding from VCs, we have to retain the data and the pitch deck for a period of time and I think it's 30 days so that the VCs you reach out to can get the pitch and get the summary of it and they will receive not only the deck, they will receive the summary analysis that you've seen on their side, a summary version. So we do, if you think about it, we have to retain it for that. Anything else? We don't retain it. And it's a good question. It says it on the side and it's something I've thought a lot about. But you know, obviously we can't. We have to retain the bare minimum. So if you share it with VCs, they have the deck to share. But, but that's all, all that we retain other than some high level aggregate data. We do have some very aggregate data. What, what are all the grades on the aggregate basis. So you can see it. But we don't retain anything that doesn't get a B or a B plus or higher. That's the minimum threshold to share to VCS that you want to share with vcs. So anything below that is gone. So anyway, here's one for data flow. It's pretty good. Listen, I know a lot of folks, are you going to say this is even better than pretty good? They're growing 73 at 48 million ARR. So it's a little bit late stage. It goes through this, what we saw before and oh, interestingly this one doesn't think we're quite ready, thinks we need to be at like 80%. We got to be minus. So I don't know if actually I agree with this analysis but it's helpful to say it says, listen, you're not quite ready to be funded. Come back later when your growth is just a little bit higher and we'll introduce you to vcs. So there's a lot more here but that is the set of tools we've built. Again, they all kind of work together. And yeah, the last one that we're continuing to work on, this one's research VCs, you can just tell us who you are and we'll tell you funds that we think are a good fit for you. I think it's pretty fun. A lot of folks use it. It's free for the Saster community. You can see how many people are using it here and I encourage everybody to try it. So hopefully that's helpful. I'm going to show you one last thing and then we can break for questions. If folks have it on just about anything. But going back to this, the other thing, this, this is small but it's been very valuable to people. This is our new agents. Here's aivc. Here's the mentor for just this can be your, your B2B therapist. It's here for you all the time. And this AI agents page, it's literally just Saster AI agents so you can find in the footer. Folks ask us so much times, guys, what do you use? Okay, and it's just all here. Here's the tools we built, the mentor, the pitch deck analyzer, the valuation, the ones we went over. But here's the third party products that these are the ones that we use. There are other great ones and we will keep adding to it. Asian Force Artisan and Qualified that we use for our outbound and inbound sales qualification. They're great. I mean literally we just closed 100k deal that qualified found for us yesterday last night, right? Amelia closed it, Qualified found it, Qualified set up the appointment, Qualified talked with the sponsor and Qualified Qualified didn't close it, but they did everything except have the call and close it. So that was last night. We use all of these. Delphi is what we use for that digital JSON Gamma for our sales decks. Momentum and attention we use to automate Rev Ops. And so our, our whole sort of suite is here. You can play with it, see them, it'll link to them and they're all here. We'll continue to add and what, what we're stealthily adding is over the next couple weeks we'll actually add all of your favorites. You can actually, this is sort of in stealth. You can commit submit agent if there's something you like, whatever AI agent you find useful, come here, submit it. If it gets approved, it'll be added and then the community can upvote the best of them and then you can even see a couple conversations where we show everything. So our whole AI agent playbooks here, I think we're pretty far ahead of a lot of folks and literally every day we get multiple questions even from public companies, big leaders, startups. What do you guys use for your Agents. So it's all here for, for you guys to see. So thanks everybody. That's the summary. Go to Saster AI and AIBC and try these tools. They're all free and fun. So with that, Emily, you think we have any questions?

There, There was a question on the live stream. I know we had it separately, but I don't know how much you want to touch on it since that's okay. A lot of this was vibe coded. Yeah, I think you've answered this, but I don't know if your answer has changed. Like if you start with vibe coding for a project like this, like what, what the question is, like what should they focus on for vibe coding?

I mean, look, here's the, here's the thing. This stuff's pretty cool. It, there's really a lot here. Like, like you know, if we just. On the AIVC stuff, this really is like these, these are all components that are being reused, but they're really five different products. Like they're pretty good. Like they're not perfect but literally we're, and I'll, you know, this is the second part of my question. They're constantly getting better. Every single day. There's improvements to these products. It is, it's, it's a lot of work. And I would say going back to the startup valuation calculator, so you know, coming up on 600,000 calculations, we probably launched that six weeks ago or something like that. So let's say, let's call it 8 weeks total invested in this, which really is not a lot to have five products in production and there's others we built. But eight weeks, let's say it's been an average of two hours a day. So that's 120 hours put into this. And that's after 100 hours of my first project blowing up and not working. So I had 100 hour learning curve and then 120 hours to get. So it's over 200 hours to get to the point to get to these tools built, which is a lot considering it's not my full time job. But I, I am much better now. Like building versions of these are much, much faster as you, I use replit. Other tools are great too, but once you get really, really good at it, it does get faster. But this is 200 hours and probably rough and tough. Was that the question?

Yeah, that was the follow up question which was how much time you spent on this and then roughly how much budget have you allocated to this?

Yeah, and so I don't know what the Budget is. I do love replit, but it. It kind of obscures how much you spent. But let's assume we've spent five or six thousand dollars on this. Maybe more than. It's not cheap like it's cheap. That's so cheap compared. Here's the thing about agents and AI. 5 or 6,000. 8,000. That is so cheap compared to a human. Where am I going to find a developer? Then I got to qualify them. Then they quit, then they don't want to work on it, then they don't finish the project. Then they do do my bugs. You can't find a dev shop to build this for eight grand. But one, it's a lot more than just Squarespace or wix site for 30 bucks a month. So like if. If that's your comparison, can I do it on Squarespace or wix? Just do it on Squarespace or wix if you can. It's easier and it's a fraction of the price. And also the thing people miss is six or eight grand's a lot. But this is my time. I mean, I don't mean to be have any hubris, but. But my time is valuable. I mean I'm investing hundreds of millions of dollars a million. I have to run saster. We have 1012,000 people come to our events. We produce 40 pieces of content a week. I don't know what my time's worth per hour, but it's a lot. Okay, so like if. I mean, what. So the soft costs of these tools, it's just really interesting if, if you've got nothing else to do, if you're. If you're between gigs or, or for me, I just like it. So I do this. Even Amelia now too. She's got it gotten kind of addictive. This is what I do instead of playing a video game or instead of screwing around or. Or doing things I shouldn't do. Like the. It's actually very efficient because it sucks up all of my spare time versus things that have low roi. But, but. And so, and so the more free time you have, I guess, the lower your soft costs. But I don't know that most people have enough time on the side to build a lot of these tools. So it better be very important to you. It better be something that you really, really want to build. And then remember, it better be something you're willing to put a hundred or two hundred hours into because no, we can do a different one where we just vibe code something from scratch. You can get something that looks like it works in an hour or two or even less, but it won't really work like you have to. I spend an hour a day fixing things, improving things, enhancing things, all across this, all the time. And if you don't have that time, you're going to be very frustrated, at least in 2025 with what you get out of it.

There's a follow up question in the chat, which is what is the average length of your prompts when you buy code, do you begin by outlining the vision or goal of the product or outcome before you start coding?

What is the length of the prompt? Right. Is that the question?

Yeah.

Well, first of all, it's a great question for folks that haven't done this before. If you use a product like Replit or Lovable. I mean, prompt is kind of a nerdy term. We all talk about prompts, right? But prompt really just is what's the, what are the inputs you give to the product, to the, to the agent? And it is interesting. These products are getting so much better more quickly. So I started on Replit in V1, now they're on V3. V3 really is, is a, a step function better? And so I think when, when it, when you start, you needed these prompts to be very prescriptive, very descriptive. Now the more you do it and the more you stay keep working in the same window with the agent, the simpler a prompt you can give. So that's a lot that I just said. Let me step back for a minute. If you haven't done this before and you've never built software before and it's actually okay, go into Google Doc and write the longest description possible of the product you want to build. Don't write it. These apps are going to encourage you to write in one sentence. I want an AI CRM for Amelia's events and all the folks coming to Europe. Build it for me. Like you could do it in one sentence, but it's going to be terrible. Okay, Instead, in plain English, spend an hour, an hour writing every single thing you want that app to do. Everything you can think about, every corner case, every niche feature, every exception. Just write it in English and then take that English and put it into Claude or ChatGPT and say, I want to give this to Repl it or Lovable as a prompt. Help me turn this into the best prompt possible. Iterate it in ChatGPT or cloud that you use every day. It will then organize it for you and then stick that into replit or Lovable to get your V01, that's the best path today, okay? For folks that haven't done it before. And I still think you should do that. Plan, plan, plan. If you can't spend an hour or two planning an app you want to build, it ain't worth building. Okay? I mean, in the old days we would spend a month doing wireframes and mockups and everything before we asked any developer to build anything. Now, you could do it in an hour or two, which is very disruptive, but you can't do it in a minute no matter what the ad said. So write it in English, put it in Claude chat, gbd, tell them you want to do a Replicator lovable build, and then give that to Replica lovable. You'll. You'll be, you'll be, you'll be far ahead. Having said that, today, these apps are much better. Replit has a very detailed planning mode, so if it's not great, it will help you work on the plan before you start. It's. That has gotten better and better. And then. Okay, this is a nerdy answer. One of the big things that's changed in replit today in v3, that didn't exist before is your context window can be open forever, almost forever, as near as I can do. So when I started on Replit, the top piece of advice was, you're going to be. The agent's going to hallucinate, it's going to get confused, it's going to create a million bugs, and what you have to do is constantly reset the agent, start all over again, start the window again. And that was okay advice when I started 100 and something days ago. But the problem is it forgets everything. It's actually terrible to reset the agent window. Like, I literally just built the calculator and I say to the agent, we just did this on the calculator. What should we do on the next project? It wouldn't remember. It was brutal, right? And it would save some things. But now you can leave this, this window going for weeks or months. It's so powerful. So now when I want to go from the valuation calculator to the benchmarking to a meet a VC thing you haven't even seen that Amelia works on to connect founders and VCs at Sasser London. The agent remembers everything. It's so much faster when the agent remembers, forget some stuff, but it remembers almost everything you've done. It's so much quicker, it's so much faster. And so that's a superpower. If you jump from app to app to app that have no connection, it doesn't help you. But if you're continuing to add stuff to your app, it's much better today to get something done, get it built, and then stay in that same agent, that same context window, and it just will remember everything. So that's why I'm much more productive today than I was before V3 and when I started. And so to folks that maybe tried these apps, you know, a couple months ago and were frustrated, I'm not saying they're perfect today. I'm not saying they meet the marketing hype, but try again. The latest version of Repl. And I'm sure it's true of Lovable and Bolt and others, but the latest version is much, much better than even 60 days ago. It's much better.

Yep. There is another question too, which is for the. Yeah. For the cash burn widget you showed.

Yes.

Can you reshare that and then what is the data that that needs to work?

I can reshow it. So it's a good question. Hold on, let me share my screen here. This it right here. Can you see it, Amelia?

Yep.

So all this is doing. If you go to. Again, if you just go to Saster AI and just click benchmarking, it'll just do this. All it's really doing. It's. It's not as clever as it looks, but it is clever and it's. I'm actually gonna make this even more clever in a couple weeks. All it's really doing is, as best it can, trying to extract information for whatever it has, your financials, your reports or whatever. It's. It's trying to figure out where your revenue is today, and it's trying to average as many months as it can get to get your growth rate. So up here, as you can see, it's a little small at the top. Last three, three months compounded. So it was able to extract three months of data. This is a classic Saster post of how you model. You take the last three or four months of growth and burn, you average them. That's most likely what the next 12 months are going to look like. So that's in essence our formula. Get three or four months of growth, that the last month was 7%, the month before 6, and the month before that 8, we're going to average that to 7% growth. If your burn was minus 100, minus 120, minus 110 the last three months, we're going to average that to minus 110. And then we just roll that forward. It Sounds simple, but I've been doing this for a long time. It's always more accurate than any cfo, VP of Finance, VP of Ops says. It's always more accurate to take the last four months of growth and burn, average them, and just roll it forward. You can, you can play with it a little bit. You can, but it's almost always more accurate than any biased human. So that's how we do it. We take that, we roll forward Your growth rate 12 months and we do it each month based on that. And we do the same thing with your burn rate, but it's super accurate. We're going to add a feature at the bottom. You'll see we hide some sneaky stuff at the bottom. ARR. Growth Planner. This is actually already built. I just haven't launched it where it takes this even further. And you can, you can, you can play with it and you can add more data if you, if you really want to go deep for your planning purposes. But it basically just takes your last couple months of growth and last couple months of burn and rolls them forward as an average. But I find that my, my, all my experience in B2B, this is pretty darn accurate.

Cool. There's another question from the live stream, which is you showed earlier that there's been, I think, like a thousand decks analyzed, but it shows that 95 are unbendable. Or you said it was giving a lot of Fs.

Yeah, I don't know if it's 95, but keep going. Yes.

Okay. How does this compare to actual VC rejection rates that you're seeing?

Hold on, I'll just show you a little fun thing. There's kind of a stealth feature here. If you go to AIBC and you click in the middle, says, dive into the data and this will actually show you all the. This is more data than you could ever imagine. I don't think that many people look at this, but it'll show you the distribution. It will show you how many are different stages. It'll show you all the average valuations and growth rates across everything. We don't store the decks, but we store the aggregate data. So you can go back here and see how it looks across verified or thousands of other analysis. It's all here. You can play with this data to your heart's content. The question was, okay, so what, what have we seen so far? 36 are Fs and 45 are Ds. Sorry, 45 are D.6%. So 42%. Don't, don't. Aren't in the zone. This C minus, D plus thing is slightly buggy. We'll actually have more of these going forward. This gap was something that probably shouldn't have been there. I mean, you know, I guess how D plus is kind of a weird grade, but anyhow, this will, this will get filled in. But basically a way to think about it is any. These are all completely unfundable. So that's 320 some odd. And then these, these Cs, these Cs and B minuses here, there's almost no chance. Right. It's just the growth has slowed too much. So the question is, is this represent what most vcc. Is that the question, Emilio?

Yep, yep.

I think this is actually better than most vcc I think. And if you talk to any investors that have any ability to submit on their website or forms, actually they get more worse decks than this. They get more Fs and Ds and Cs than this. I would say 90% of the of the decks I've been sent over the last 10 years are C's or worse. And. But the thing is, it's a weird metric because you, you learn to just ignore them. Like the deck is terrible. Like you can see that. Like Dear sir or Madam, I am, you know. Okay, well I don't. I can open that one but there's really no need when it says Dear sir or madam. Right. Or you can already tell it's abysmal or it's an investment bank or there's just a million problems. So I think those folks, the dear sir and Madams and the investment banks and the scam artists and the LinkedIn messengers, they're not even using our tools. So I think this actually flatters the distribution. If we got every deck, it would, it would look worse.

Do you related question. Do you know yet if from the A's and I think you said the B plus is you save the decks. Do you know of any funding that's come out of it yet?

I've done one last week that hasn't been announced.

There you go.

It might be the only one. I know a lot of meetings have happened. I think it's the right question. I think the truth is ask us in about 90 days. There's going to be a lot more coming up. Disaster AI London, the this. I mean honestly, the pitch deck analyzer has been out for a while. Our VC, our VC going back here, our VC matchmaking. We haven't even done any PR around it. It's made 170 intros so probably, you know, it's, it's a, it's, it's a small number of investments but it's a lot of good meetings, which matters and we'll know and. Well, no, in 90 days I, I think it will just get better and better. What I can tell you this, I think, I think it's the right question. I think what I can tell you, you know, 45 days into this experiment, which is a lot of usage, 1833 decks, 169 intros. What I can tell you is that it works. And what I mean is when a really good deck goes through that is matched to the right vc, it gets opened immediately. It gets, and I, I get emails from vcs and when we've made little hiccups like we, we have over graded decks and we've probably undergraded them. I mean, you know, for not reading every deck manually. You know, I try to QA it, okay, but I did, I can't QA18 33 decks. I can't even see them right. But, but I, I've QA'd my own decks. I've QA'd things people have sent to me. It's pretty good, but it's not perfect. But my point is when we've, we, when, when we've made some extraction errors and sent it to vcs and they were claim to be better than they were, boy, I get my inbox lit up like why'd you, why'd you, why'd that damn AI send me the wrong thing? And a couple times and I've gotten, and I've gotten one feedback from a top vc, which is this one was so close to what I want to do. So my point is people read this stuff which is the most important thing and we're going to keep actually iterating it so that your pitch is better and better. But the most important thing is to the learning from this is listen, you can reach out to VCs yourself. You don't need this tool. I've, I funded multiple unicorns off cold emails, but I find most founders aren't very good at it. They don't know the right deck, they don't get the right feedback, they don't need to reach people. And we're only onboarding VCs that no one trusts Aster. So it's, it's a, it's an asset when it comes from us versus some random person. And you know, fundraising isn't an equal ground. And it does help if you came out of Stripe or Y Combinator. And the real point of this tool is not to help everyone going through this batch at Y Combinator, it's to help all the other great ones that aren't out of Stripe or Y Combinator. And so take advantage of the fact that Saster does have a brand, it does have a channel, it does have a connection that folks from Cowboy and Altos and USVP and Scale want to be part of this program. We can't get you funded, but if your startup's really good, we can actually get you in front of people. And that's, that's what has historically been hard.

Yeah, there's a few others. One. Another one was if I'm a founder and I rated low, what should I do?

What should you do? Yeah, well, look it, we, we rate the team and the deck. We don't really rate individual founders per se yet. I mean that's part of what we do. But there's only so much we can do to, to go deep on the founders. We do a little bit of it. Listen, you know the answers. If, if you have no pedigree and no traction, it's hard to get funded. This has always been true. No revenue. No one's heard of you. No one knows you. That's the toughest of the two by two. Right? You know, if you just, if you just. If you have got a, a crack team out of open AI and you've got a lot of revenue traction, you'll have infinite funding. Right? That's the other end of the two by two. So what is the answer? The answer is the same as it's always been. It's just our tools will keep you honest. Either recruit a world class team, which is hard, but everything's hard. Everything's hard in startups. Recruit that team. That's your job anyways. A founder is to recruit or the practical answer is you're going to have to get more revenue to get funded again. Three folks out of open air stripe, they're going to get funded without any revenue. Three folks no one's ever heard of outside of any tech ecosystem with no connections that no one's going to vouch for and no revenue. There's not that many people that just invest in vision with no proof at all of the founders. So get, get customers. Revenue solves all issues and even just in the early stage, a little revenue. But fiber growth solves all issues.

There's another question in the chat. Under detailed metrics by stage. You know where this is?

Okay, I don't Know if I know.

Yeah. Keep going down. Yeah, yeah. Okay. On this one, can you clarify if the series A metrics are after raising series A or in order to effectively raise the series A.

After series air in order to. It's a good question. The. The it's there is the man's. I don't know that it matters. A huge difference. The valuation calculator that's been used 500,000 times, that is based on closed rounds. Okay. So that represents deals that have happened. To answer your question, this is based on the data that we have. So it's, it is what we think it takes to get funded based on that. But these are not closed deals. These are deals that have been run through our system. So I guess that sense this data is slightly worse because this is what all the Data, including the 5,000 closed deals, tells us. But we're not actually getting those deals back from founders and then rerunning them through the system. But I'm not sure there's like a huge difference, but. But that's the answer.

Yeah. Related. Are there benchmarks you have from the data that would show like, what you. What the metric should be for each of the different series?

I mean, there, there absolutely is. I mean, literally, if you go in and if you just run, there's different ways to get there. But if you do any of the tools, the benchmark, the pitch deck analysis, the benchmarking or anything, you'll just get it. You can click on here and find out what the, what, what these iconic benchmarks are to actually get funded. You'll see them by stage 10 million below 10 to 25. But even more, you can scroll down. It's. It's really here almost ad nauseam. You. Here's benchmarks from Carta, Bessemer, Saster and Iconic. You can click into all of them. You can see all the details. Click on this little eye, you'll see more of a modal information on how they're all built. We can add even more. But if there. There's four detailed sets of benchmarking with a lot of detail on how they work right here at the bottom.

Okay, there's a couple more and then we can break for the ones that have scored higher.

Yeah.

Aside from the numbers, is there anything else that you see as a trend in the decks that would optimize for a great pitch deck?

Yeah, the number one thing I've seen that founders have gotten wrong so far is they don't put enough in the deck. The. At the end of the day, this is run by AI and we do multiple passes through Claude and then we run it through some heuristics. So there's sort of three passes going through this data. So it's. I think this is better than anyone else has built. But it is far from perfect. It is very, very far from perfect. And that's why I have to like work on it every day. It's kind of fun and kind of stressful. Not stressful, annoying. It's like a jigsaw puzzle. The number one mistake I see is founders uploading. Too simplistic of a deck with not enough data, not enough background to really help VCs understand it. Now listen, no one really wants 100 page deck. Personally I actually don't think there's a problem with a long deck as long as you like have an intro portion that's 15 or 20 pages and then you have like an appendix with more. I think you can have a 40 page deck as long as you've got a main and appendix. But people just don't put enough in where. Because the problem with VCS is this will do a really good job of helping you tell if it's, if it's of interest. Is this something I should spend some more time in? And it's especially valuable again if you're not in the inner circle. If you're not from YC or Stripe or OpenAI, it will get you a look you might not otherwise get objective speaking. But then sometimes the decks just, there's just not enough meat there to get people excited. So the mistake is like, and actually now that you say it, I'm going to add more and more to this. We can almost add an excitement factor. We have other factors. Right. But I'm going to encourage people to add more to the deck. Like and this is so great, it'll score it for you, it'll show your weaknesses, it'll show you everything. But then ask yourself and ask folks you know, don't hold back, don't like, don't do teaser decks, don't do coffee meeting decks. Like you don't know these people. Put, this is sales, this is selling stock. It's not selling software, but it's still selling. Put your best foot forward. Right. How many times have you taken a sales meeting for a vendor and you've shown up and it's. There's just not enough information to decide if you want to buy the product. You just don't buy the product. So rambly way of saying put more, not less into your deck.

Yep.

No Teasers. Don't tease people. Teasing works. You know, it works great. Again, sorry to be redundant but people, this is one of the things. There's just such terrible advice on the Internet if, if I get an email. Hey, we're three guys from open AI. We're the top three engineers in this part of open AI were dropped out to do a startup. You want to take a meeting, of course you're going to get that meeting. Okay. Those guys can tease. You don't tease unless you're sure. You can tease. Don't tease unless you're sure. Put it all in there over. Describe why it's great. Make sure that you would not just like in, in sales for B2B. Make sure you would buy your own product based on your collateral. Make sure you would invest in your own startup. Make sure you invest in your own startup. And if there's anything good not in there, put it in man. Put it in.

Yeah, there's a few comments on the chat on like, like technical, technical requirements for a great pitch deck. I guess one that would rate high in these terms of.

Yeah.

Have you seen either an optimum length and then does design matter?

I, I don't know if there's an. I'm taking mental notes of this. We'll write down because we could add some of this. I don't know that option length isn't something we've tested. There's a lot of stuff we've tested. The lack of depth isn't is issue for, for the decks design we have. I mean that, that I can tell you because the ones that I put through are beautifully designed. On Gamma, we don't, we don't test for aesthetics at the moment. May, maybe we should. I, I, there is controversy on whether that matters. Certainly when I look back the best investments I've made when I look back on them, they're seed decks were rough but I think a crummy looking deck in this day and age is really a bad sign of lack of attention to detail. But we don't, we don't really judge aesthetics. We have this deck quality score you can click into the modal. It's got team quality, market size, traction, competitive advantage in your deck, your growth strategy and your market heat. So these are what we do if you click in which is a lot to do but design and length are not yet a criteria.

Yep. Related. Maybe last is question. Is there a tool or have you thought about adding a tool here that would help you workshop your deck if it doesn't rate high?

We should we have a placeholder here? It's actually literally here. We. I need to do more one. It's called your pitch deck. So if you sign up for this, we send you a little bit of a sequence around it. But we've just begun doing it. We've, we've. This is, this is early. It is here. It's a good idea. So sign up here, you can fill it out and then we'll add you to this track. It says join the wait list, but it's kind of already started, so. So we're aligned on this. In practice, we have seen folks just upload their deck multiple times. Like they'll do it once, then they don't like it, they upload it again. They're like, okay, maybe Saster's right. And then they come back later in the week with a better and better deck. We have seen multiple folks inch their grades up and I encourage you to do that. That's what it's here for. So that's the best way to do it is there's tons of feedback here and I'm not logged in, but you can log into this full investment report and you'll see even more feedback. We'll tell you especially here what to fix and what to improve. Do it, upload it and see if you get a better score. But we are going to plan to add more for folks that are sort of. And again, don't take a person their tough grades, but C plus to Bs, we are going to add much more here. You can just sign up and do it and we do a little bit, but more to come.

And does this tie in the JSON AI on the improve your vision?

It doesn't. It doesn't. And there is, it's a good, you know, it's a good question. The reality is all of this stuff intentionally, we built on this one platform, right? It's all built on Saster AI. Saster AI has a lot of stuff. It has news, it has market data, it's got our VC. But the, the digital JSON's built on this Delphi platform that doesn't even have it deprecated. Its API doesn't even have an API anym. So it is if that, that data is kind of an island. I wish it wasn't. And maybe someday we'll replace it with another vendor or they will return their API. But right now it's, it's. Everything else will be integrated here, but this, that one's kind of an island.

Cool. Well, we got some questions on the agents, but I know we're going to hit that next week, so we could.

Do two minutes before we break. It's up to you. You're the boss.

We've done a lot of them. Yeah, let's. Yeah, we're.

Do we have one more question on this stuff or on this stuff?

No, I think you've hit those. Yeah, there's some questions on the agents, but I think we're going to do.

Amelia's going to do a deep dive on agents, I think next week. Right. Not only we're going to dive, we're actually going to show you how we do all. All of our three core SDRs, BDRs, and that branch. You're going to show you behind the scenes. We're going to do it for real. You're going to see it all. It's going to be great. So come next week and bring those questions. And that's the whole point of next week. We'll do all the deep dives on it, especially the out the. The sgr BDR stuff. Right. That's. We're going to focus on next week.

Yep. Yeah. Because some folks were asking us about, like, what results have we seen on those agents, which I have for next week. So I'll show you guys them.

Cool. All right, thanks, everybody. We'll see you next week. And we'll see you in. Don't forget to come to London. It's mandatory. December 1st and 2nd. We're gonna. And if you want to talk agents, we're gonna show you everything live. You're gonna be able to talk to the folks that build them. You're gonna be able to talk to us. We're gonna have 2,000 folks. It's worth a quick trip over to London. Christmas in London will be great. We'll see you guys. See you next week. Thank you.

Sounds good.