SaaStr 817: What Happens When AI Kills Your SaaS Growth? Live Q&A with SaaStr CEO and Founder Jason Lemkin

0:01Welcome to the official Sastra podcast where you can hear some of the best Sastra speakers. This is what the cloud means. Up today on the Sastra Podcast. I've dropped everything. I'm here for two months just to learn. What do I do? First of all, if nothing else in SF, there's endless meetups, hackathons. Like, just ingratiate yourself in it so that the stress that Aaron's feeling, that Yamani's feeling, that you're feeling, that you're just acclimated to, it's a new normal. I know it sounds silly, but it's not. Like, adjust to the stress, right? Adjust to the pace. Adjust to the 2x pace. even me, like it's not the same, but even I'm working in terms of output, I'm working twice as hard me as a year ago.

0:45And then, I mean, this is only sort of helpful, right? I'm just trying to be a therapist. Meet all these people, hang out with more folks that are stressed out, right? But even if you don't really care about the content, socialize with the people that are stressed out, right? And then figure out a way to work twice as hard. Ship features twice as hard, Make the product twice as good. Hey, everybody. Sastr Annual will be back May 2026, the world's largest SaaS and AI gathering for executives. Just as last May, we hosted 10 ,000 attendees with 68 VP level and above attendees, 36 % CEOs and founders, and 25 % were AI first professionals.

1:21It's the very best of S-tier attendees and decision makers that come to Sastr Annual and AI Summit each and every year. But here's the reality, folks. The longer you wait, the higher ticket prices get. They're cheap now. They're cheap. So just get them. Early lock in your spot today. Use my code Jason100 for exclusive savings. Get your tickets at podcast.sasteranual.com or just use code Jason100 when you check out. See you there. Saster Annual and AI Summit 2026. It will rock.

1:51Thanks everybody for being here. We just have a tradition. This was a pretty good, everyone have a pretty good day three? Yeah. So we don't have to do it, but usually we've had a slightly quieter day three this year, which was probably our best. So I sort of did a tradition of it, trying to answer any questions we didn't get to over these days. But I will say in closing, I am really happy this year. I will tell you, it took me a while to reflect. But how many folks were here last year? Yeah. I mean, last year, you probably thought it was good. I actually thought it was the worst Saster ever. It had nothing to do.

2:24The production values were great. It looked good. It was well done. But in a million hour agreement, the vibe sucked. The vibe sucked. And it sucked because people thought we were in a downturn. They thought AI sucked. There was a lot of woe is me. And I'd walk out and meet founders last year or executives and half would be like, man, everything's terrible. I can't close anything. And it really weighed on me the last 11.9, two months. And man, the whole three days, it's been great. I haven't had a single Debbie Downer conversation the whole time it is, right? And if you were here with Aaron Levy, he's like, I'm stressed, but excited.

2:59That's how you should be. But yeah, it's hard, but you should be excited. So for what it's worth, I want to take a question, but I'm reinvigorated with AI. I'm excited. You can see it in what I've done. And even just the tenure here, there's like the energy is so much more fun than last year. Last year, I kind of had to put the smile on my face after I'm like, we did it. It was great. But man, it's like, uh, there's a Debbie Downer vibe in B2B and it's behind us. So like, I'm super excited. Not a single person I met here said, oh, we're in a downturn. life is terrible. I didn't. And thank God.

3:28I think they all stayed home this year. And I think they should either go get excited about AI and join something new or just stay home. I don't know which one, but thank you, everyone. It's been a delight for me and Amelia and the team this year. Really, really great. And total 180 from last year. So thank you.

3:51So I have two thank yous and one question. first of all thank you for uh doing sass for in may uh this time around it was my birthday yesterday so this was like happy birthday birthday gift yeah second thank you is three years ago roughly it was in september my co-founder abdul asked you if i remember why c is uh worth doing yeah we took your advice we did it it was the best thing that you know happened to our company so good and i think as the demo uh the ai demo has just shown why c is still very much worth doing in 2025 now Well, you guys are perfect. You guys were for what it's, I mean, ask your question, but I'll answer the different question.

4:24Why I told you to do that. I mean, I do believe in YC, right? But you guys were the perfect, you guys had a ton of energy, a ton of drive. You knew a niche market, but you were outsiders. Nobody knew you. Right. And so I'm not saying everyone should do it, but anyone like that, that goes through YC, it is a gift, right? You'll be validated. It's, it doesn't, it doesn't get everybody customers. It's not going to get to even if you're in solar, it's not going to matter. But to take someone that didn't go to MIT or Stanford and wasn't the top engineer at Stripe and didn't work in the... It really can be magical, right?

5:00And probably more so today, right? But that's why I said it to you, because I'm like, you guys have the raw ingredients of success, but this will just lever you up. Awesome. And so the question is, as you know, we're a vertical SaaS, That's very much like a traditional layer K, just now starting to think about AI. In the best possible universe, just because these are our laws of physics, small market, we're going to be triple, double, doubling. In the best possible universe, we're never going to have a windsurf or a cursor style J-curve. So here's the question. Listen to your chat with Rory and then Harry a couple of weeks ago.

5:37So here's the question. Is that still kind of an attractive asset class, right? for maybe 1.0, 2.0 investors like yourself who understand just how rare even triple, triple, double, doubling is and how do we just conceptualize this as a vertical SaaS company, first, AI, second in today's day? Well, it's a good question. I think Rory and I both have, and we did address it. I've talked about it a couple of times, but it's so important. Let me restate it, right? Which is, look, you're right. Until 18 months ago, for folks that have been in B2B since Aaron and I started, the gold standard was triple, triple, double, right?

6:15It was triple, double, double. However long it took you to get to a million, one week, 10 years, no one cared. Took UI path 10 years to get to a million, right? There are plenty of, took Squarespace like five years to get to a million in revenue. They all did well. Within limits, no one cared, but when you got to a million, they wanted to at least see you triple, double, double, one to three to nine to 18 to 36. That was the path. And then what I actually learned from Rory on the podcast, The second time I met him was at an upfront summit in LA in 2013 with Mark's sister. And he's like, now the new bar is one to 10 and five quarters or less, the best ones.

6:49And so I started thinking that was the prime version of triple level. I didn't know until Rory explained it, right? Because I didn't know anything. But if you do the quick math, it got you to a hundred million in an IPO. Triple, triple level got you to a hundred million before your grandchildren were running your company. Okay. When Rory invested in Box with Aaron, when Rory invested in HubSpot, those are pretty good ones. Shopify. These guys all IPO'd around$100 million in revenue, growing really fast, often 50%, 60%, 80 % at like 800 million valuations. That was how it worked. So you're like, okay, I got to get to$100 million.

7:24And people would IPO in seven to eight years back in the day. Now they're going to IPO, like eToro just IPO'd today. I think they're like 16 years old or 18 years old, but you're not at 100 million. Now they're all at 500 million. That was last year. All the IPOs that we've had in B2B recently, Klaviyo, OneStream, Rubrik that was here, they're all 500 million growing 50%. So then stop your math. Okay. Now I got to go even further and faster. Then the new math became 550, not like 50 or 60 to 100. Then AI changed it all. And everyone's like, holy crap. Some of these startups, for whatever reason, are growing at insane rates.

8:05And then what I started to see about a year and a half ago was a lot of the B2B investors I knew old school, the bar had just gone up. Right. And like, I had like my catch up with Byron Dieter, who's OG, you know, was first one of the first service type. So many great ones. And I met him about a year and a half ago. it was like yeah honestly we're only doing ai investing because the bar is so high he's like i'm that's it right and then then they went on to invest in anthropic late and others and that's what they're looking for and so when someone like byron that is as og is that uh you know been a sas ceo himself done box done so many early ones is now looking all in on this that's where like 80 % of the B2B investors are.

8:52Now, if you listen to Rory and I both said, we still like triple, double, doubles. So what is the long-winded or longer, less TLDR version? If you're triple, triple, double, double, as hard as that is, as hard as that is, maybe only 20 % to 30 % of the investors that would have been interested 18 months ago today, 20%, 20%, 30%. So you can say, the woe is me, or you can do what you have to do in sales. Talk to more prospects, talk to everybody. That's just the reality, right? But the only thing I would add at the end with that is the dagger. And you probably saw it with Jack from Clio and Tidemark, if you were there.

9:32Overall, there's a little more in vertical SaaS today than in other categories. So you'll get more meetings. And it's viewed as more durable. Most vertical SaaS companies had no downturn, so-called downturn. I was with the CRO of owner this morning. They had no downturn in restaurant SaaS, right? So there's a lot of reasons people are more interested in vertical SaaS, but generally the investors are taking more risk for more growth. And I know it sounds crazy, but if you listen to what I did with Rory and Harry, a lot of investors, when Fox and HubSpot and Shopify IPO, they're looking for billion-dollar outcomes.

10:12okay then if you came to saster 2016 2017 everyone's talking about decacorns 10 billion outcomes right if you listen to that 20 vc we did vcs are now looking for trillion dollar outcomes people have decided that every couple years there'll be another open ai not every week there'll be a trillion dollar company like these companies are smaller facebook wasn't trillion google wasn't trillion tesla wasn't trillion spacex isn't trillion now that there's multiple trillion dollar companies, it's no longer viewed as an anomaly. It's viewed as a factory. Like we will create, and folks believe that AI will create more of them, right?

10:50So if you're a VC and you have a lot of money and you want to get rich, do you want a 2 billion, a little 2 billion dollar exit? Like Databricks just bought this, some Postgres company for a billion, right? Does that really make a difference for your 2 billion dollar fund if you own 10 %? 10 % I made 100 million, hooray. So a lot, if your fund is 10 million, if your fund is 5 billion, then a hundred million you made off that is 2%, right? 2%, right? Doesn't even go above the line. So this trillion dollar outcome, guys, it's nutso. And you can laugh at it and you don't have to take it too seriously, but realize smart people aren't laughing.

11:31Like they're really, the Peter Thiel's and Mark Andreessen's are like, how can we find the next trillion dollar company? Trillion dollar company, right? And it probably won't be you, but maybe it is. Like, I'm not saying it won't be, right? No more questions this year. Just thank you again. You're very much the Rory of my generation and hashtag CryMeRiver. Thanks for being back. Congrats to CryMeRiver. Yeah. How you doing, Jason? So we are a three-year vertical SaaS company focused on energy. When we started the company, the first two years, it was a SaaSter playbook and we literally followed it religiously.

12:07And it worked for those first 24 months. At least you got 24 months. Well, there you go. And so in the last 12 months, it started breaking down, right? And I think you'd be the first to kind of admit from your comments today and during the week that a lot of the kind of traditional playbook is shifting in this kind of world of vertical AI agentic. I do think, keep going, not to interrupt, but just be, I do, I want to be clear what I'm saying because sometimes it's, I actually think the traditional playbook act, not with saying what Kyle and I talked about this morning. If you have intense product market fit, it works fine.

12:37Yeah. It works fine. OpenAI is doing field events. OpenAI is doing webinars. OpenAI has swag. But OpenAI has so much product market fit, right? The playbook works, but when you are falling out of product market fit, it decays rapidly today. Yeah, well, it's more, you have these canonical pieces of content, right? When to hire the VP of sales and so on. The context of what you just said, do these need updating? you know are these kind of core canonical pieces of you know content about how to build your company how to get to that first one you know zero to one one to ten etc do they shift in this kind of new like area of i don't think it shifts i mean in all in all seriousness if you read saster carefully you'll actually see i am updating a lot of the canonical pieces a couple times a week i'll take some of the really good ones and i rewrite them and update them so i'm cognizant of it don't get me wrong but if you watch what i said i think 30 to 40 percent of inside sales will become ai i think we won't need mediocre customer success.

13:38I think marketing is changing. I think it's all happening, right? But the motions are still there. So even though Kyle and I had a little bit of a debate, I don't challenge people, a debate on whether a CRO will manage AIs and humans together, I actually think he agrees when he thinks about it. Yes, but it's still the motion, right? It's still discovery, demo, solutions solving. Even Streetart from Snowflake said it's the same thing. It's just the speed and the quality are changing. We all need to know how the systems, the instrumentation, the quality changes. But even if 40 % of sales is done by AI, we're still having sales.

14:16I have no problem with it all if million-dollar snowflake deals can be done by AI. I think it's cool, but I don't think anyone's saying that. I just think it's more you just have to run these playbooks with AI, with intensity, with curiosity, and everything is faster. Maybe the part that is the most dated of that playbook isn't really dated, but it feels dated, which is we just have to move so much faster, right? We just can't get back to a company in a week and pretend we have three competitors. We just can't. And I honestly think if you work twice as hard and four times smarter than last year, it all works.

14:56But most people are not working twice as hard as four times. Actually, the horrible truth is most people are working less hard than last year because they're bummed, because it's not working, because there's a downward death. If you don't think the playbook works, you invest less in it. You don't take it seriously. And you just have a, I see this downward spiral. Sales isn't working as well. So I don't try as hard, right? I just want quota relief. I just want to sell less, right? And then it becomes a self-perpetuating prophecy, right? I'm still learning. I'm back in student mode. I'm updating everything.

15:28I think sales, marketing, onboarding, support, success, they're not going away. I hope more and more of it is AI, and I think it's under radical change, but the underlying motions will be the same. We still need defined apps. There's still too many apps. AI just makes there even more apps. We're still looking for brands. I mean, one of the top questions I got at this conference is, which ai tool should i use for sales okay all over twitter all over everyone get you got kyle's answer you got my answer why are people asking that because there's no dominant brand right we need 80 of folks buy with brands right and so brand still works and so the irony of this all of this is as much as sales and marketing changes probably the one thing that won't change is brand marketing meeting folks in person that you've only met on the internet if you listen to folks that are frustrated with go-to-market on the internet they're like we've got to do events in the field because nothing else works right there's a bit of a negative i think that's overstating it but my point is brand and multi-touch and reinforcement is going to work in this is digital age too but um it's got to be faster it's got to be better thank you yeah hey so on the new playbook that you're talking about evolved um the faster ai infused old playbook yeah yeah yeah so what do you recommend your top three recommendations to go faster look i next year i'll have concrete answers i'm still learning right ai cannot make someone buy a product they don't want to buy ai cannot build a brand but all i would do today if i if i're tiny i would just want to make sure that everyone that touches my product at scale has an amazing experience.

17:13Again, if we go back to the Sastra AI, 150 ,000 chats, if you came to sastra.com or sastra.com 60 days ago, you'd have no experience. All you could do is read our content or buy a ticket on the two websites. Today, we've had 150 ,000 people have had, most of them have had a great experience with us. So make sure whatever they are, however they touch you, use whatever AI and other tools you can so that everyone instantly has an elegant experience. They instantly get the information they need. They instantly get the data. They instantly talk to a synthetic human and then a human. That going, that instantly, that you don't just have a meeting that everyone's followed up with perfectly and properly multiple times afterwards and their needs are served and all the onboarding is perfect.

17:55And the handful of customers you do get, you don't forget about after and they're perfectly onboarded. And the remarketing and the retargeting is perfect. Use all these tools. and there's great AI sales tools here. Can an AI SDR bring you trillions of dollars of business? No, but can an AI SDR help you build the prospect list, target it and write better emails that then you have to manage? Yes. So use all the instrumentation tools, know its limits and make sure everyone is treated like a real-time king and queen. Real-time king and queen is my. Old Sastra post is customers tiny, small, medium and large should be treated as kings and queens.

18:31My evolution today is real-time kings and queens. Use AI so that they're instantly feel like kings and queens. That always works. So thank you. Yeah. Hey, Jason. Beneath here. It's my first time at Sastra. Loving it. Thank you for coming. And I am so petrified of what AI is doing and how I'm not ready for it. Are you more petrified after this? No, I'm just as petrified. The same petrification is when you got here. Okay. But I've dropped everything. I'm here for two months just to learn. What do I do? listen it's good first of all if nothing else in sf there's endless meetups hackathons leftathons mcpathons like just ingratiate yourself in it so that you the stress that aaron's feeling that yamani's feeling that you're feeling that you're just acclimated to it's a new normal so just let i mean i know it sounds silly but it's not like adjust to the stress right it is the new normal adjust to the pace, adjust to the two X pace.

19:31Even me, like it's not the same, but even I'm working in terms of output and Amelia too, we're working twice as hard as 12 months ago. Right. And I don't, I have my own stresses, but listen, your life's more stressful than mine. I'm working twice as hard me as a year ago, adjust to it, adjust it or quit. Right. And I know you're not going to quit cause you're here, but adjust or quit. That's the best advice I can. Right. Right. And then, I mean, this is only sort of helpful, right? I'm just trying to be a therapist. Meet all these people, hang out with more folks that are stressed out, right? If you want an SF, if you're ever doing this, you literally can go to a hackathon or an AI meetup every fucking night, right?

20:07There's just almost too many, right? But even if you don't really care about the content, socialize with the people that are stressed out, right? And not so much that you want to jump off the roof. And then figure out a way to work twice as hard. Just to clarify, maybe the stress, you know, it's not a therapeutic question. I'm asking from a playbook perspective as well. Yeah, we'll figure out how to work twice as hard. On what? All of it. The ship features twice as hard. Make the product twice as good. Work seven days a week. I mean, listen, it's brutal. I don't know. This was toxic to talk about 12 months ago.

20:44Now it's okay. Like when Kyle from Owner was here and they're coming up on 50 million, growing 10 % a month, this trades at a billion. That company works, They work seven days a week in the office. Now they've made some exceptions as they've grown. They're partly remote folks have more families. So it's okay to work six and a half days a week in the office. Adam's going to get a little mad. I mean, they're more flexible, but not really. And this, this is of all the companies I've invested in. And I have some really good ones. I think owner is the most, has the highest rate of the highest rate of software output, the highest rate of development.

21:18And the team is great. and the engineering team is as good as kyle okay on the revenue side but they don't work 40 hours a week man they work especially adam the ceo i literally every ceo works hard right come on everyone but i've adam 11 a.m 6 a.m on sunday whatever it's in burlingame it's in sf it's in florida the customer is in in orange i like i saw him the other law in orange county is meeting with the restaurant down the street when I was orange, just it's a lot, man. I don't know if this is answering your question, but you got to work twice as hard for maybe 50 % more hours. And then my last point, and again, I know my, my helpful, but if you listen to some of the things folks have said, and it may not help you, but if you have a team, just make sure everyone is using tools every day.

22:13it is good advice i mean even even amelia who runs astro with me now we use so much ai across like this logo right i mean we have our designer here who's one of the best i've ever worked with but this was done by ai almost all the drafts of this were done by ai our videos are done this app called higgs field right um to scripts ai does our podcast now and so now we do a lot more podcasts and they're a lot better right there's so many ai writes like this session that we just did here These will all get written up really well by RAI, and they'll be much better than a human. So make sure everyone on your team is spending hours a day, again, not just a recipe in Claude.

22:50I don't know if it's helpful, but it's the best I could do for today. But thank you. Yeah. You said yesterday that there are all these 100 billion revenue companies that have suddenly hit a wall. They're going to 20%. Hundreds. And quote, AI is maiming them. Yep. So can you dimension that maiming a little more? Or is it more competitors moving faster? Yes. Is it internal build? Yes. Versus build versus buy? Or is it the internal development process speeding up? Why are they all slowed down? Yeah. What's driving this contact? What's the third environment? It's probably compounding effects. First of all, one of the biggest, there's probably two issues.

23:30The first two have, there's probably a bunch, but the first two have nothing to do with AI. One is too many didn't go multi-product. If you listen to Tope, I don't know if you're with Tope for Calendly. He's not telling you what's coming this year, but when you heard him say multi-product, they're well over 200 million. I don't know where they are. I don't know if they're past 300, but they're well over 200 million. Multi-product is the game. Okay. Too many folks got to 100 million with one product. Too many folks got to 100 million pretending they had a second product, but it was really an extension of the feature, right?

24:05There are some folks like Parker Conrad that disagree with me, but although we'll do something with him on it soon. I don't actually think we disagree, but I think if your second product is smaller than your first, you're not multi-product because it doesn't get you anywhere. If you get to a hundred million and it takes you two years to launch a product that can only do 50 million, it doesn't get you anywhere, right? It's got to be bigger. So too many folks didn't go multi-product, pretended it was multi-product. It was just a second feature. It wasn't a whole new product, right? So that's root cause number one of the slowdown at a hundred million.

24:33Root cause number two. And so the lesson learned, and it was funny, I was re-watching the chief product officer of Samsara, which is one of the most top 10, he said last year at Sastr, you got to go multi-product by 10 million in revenue. You got to go multi-product by 10 million. And I think it's the right way to think about it. Okay. Second reason they're decelerating, which is they just couldn't recover from how easy it was in 2021. They couldn't recover culturally. They couldn't recover the fact that in 2021, anyone would buy their product and they couldn't recover that because of that, their product is the same as 2021.

25:09And so in 2021, they grew 70%. And then in 2022, they grew 40%. And then in 2023, oh shit, it's 20%. And then they hit this year, what did it decelerate to? So many folks went whatever that, some version, 70, 40. And then 40, the investors are like, oh, we're nervous, but 40 is still pretty good at 100, right? And then you see the 20s and you're like, well, as long as Jason turns it around to 20%, and then you fall into the teens and everyone gives up. and then where do you go from the teens you end up like just ripping off your customers because there's and then what happens after the teens is you start getting investor update and it's no new customers the customer account is the same but they raise prices again they raise prices again right so living in 2021 killed people not going multi-product killed people that alone led to this and then the ai competitors so fierce because not only is all the discretionary budget going into ai which you heard from everybody, that's bad enough that CIOs especially want to put money.

26:13CIOs want to do two things. We ran out of time in the opener. They want fewer vendors, and even Topes said that's happening in Calendly. We want to consolidate vendors, but we're putting more money into AI. So you're either benefiting or getting screwed by that. Either you're getting consolidated out of existence or you're benefiting from the AI. So too many folks are getting consolidated away, right? That's the third trend. And then the fourth one, to answer your question longly, and then this is it, is look, even if none of that is an issue, if AI just creates, if you just got five new competitors here and they each have 1 % market share and you lose 5 % market share, that even 5 % is really stressful.