How the smartest founders are winning in AI

- June 4, 2025 (10 months ago) • 01:15:40

Transcript

Start Time	Speaker	Text
00:00	Jason Lemkin	Now, zero to one is very likely in AI. You build an incredible tool that no one else can, and you can make $1 million in a week or two.
00:12	Shaan Puri	Okay. For this episode, what I'm thinking it's about is—guys, this is going to be generous to us—but what is the smart money doing with AI? In betting, and sports betting in Las Vegas, you always kind of want to watch what the sharps are doing. They tend to be maybe 10 or 15% earlier to understanding or recognizing things. Similarly, Jason, I wanted you on because you've been doing really interesting stuff with AI**: investing, talking to startups that are using it, reading all the reports, and you yourself built this AI tool. I want to get your reaction to a few of the different things. Can you first explain the AI experiment you did where you basically made a body double of yourself out of AI? What was the experiment, and what did you learn?
00:58	Jason Lemkin	Well, yeah. Saster has a lot of content. As you guys know, we have 12 years of content. Some of it’s out of date, right? For sure.
01:04	Sam Parr	SaaStr: Being a previous founder who's exited multiple software companies, you're now like the godfather of SaaS because you have this blog and media company and events business called SaaStr, and you have a huge fund. So that's...
01:16	Jason Lemkin	Kind of the most interesting part—for seeing the future a little bit—is that I have about 10,000 pieces of content I've written over the years. A lot of it is not out of date, and I update it personally. So, about 10,000 pieces of content. We also have several thousand interviews, like you guys do, with so many great folks: early ones with Dharmesh; middle ones with Dharmesh—Dharmesh when HubSpot had just IPO'd; Dharmesh the other day; Yamini last week; Brian Halligan over the years. We have all of it. Our AI ingested everything: everything I've written, every tweet I've ever done, every LinkedIn post, every SaaStr video from every speaker for twelve years—Aaron Levie the week he went IPO for the first SaaStr Annual, up to last week when we were talking about AI. All of it was ingested into this AI. What I instantly saw was that as soon as I would talk to this AI, it was much better than me. It was much better than me because I forget—how did you... [sentence trails off]
02:11	Sam Parr	Get that information in there.
02:12	Jason Lemkin	You just hook up RSS, right? The feed for a blog. It can scrape the web — it's pretty easy. Soon we'll be able to do other things like mcb [unclear], but it's just scraping APIs and RSS. It's really just scraping. You just give a URL.
02:25	Shaan Puri	But did you use, like, Delphi, or did you build your own tool? What did you do?
02:28	Jason Lemkin	I tried a couple things, and then I used Delphi. At first, we had so much content it was a little difficult to ingest. There is a little bit of work, and I spend about ten minutes a day training it, which we could talk about. So there is a bit of maintenance involved, but it's better than me. The thing I didn't get until I saw something that was hyper-trained: this is a lot of content — 20 million words — very few mistakes, very few hallucinations, and the ability to connect things that I can't. I just don't remember what Sam and I said on "The Hustle" on stage years ago versus when we did an MFM live versus today, but the AI knows it all.
03:08	Shaan Puri	By the way, 20 million (20,000,000) words is like 200 books’ worth of content. An average book, I think, is about 100,000 words. So 20 million words — you’re saying you basically gave it 200 books of your own brain and the brains of other guests that have come on your show, yes? And your goal was to build — was it to build basically an advisor, an AI version of you that could be an advisor to a founder who wants to come and talk to you and ask questions, but you don't have enough time in the day to book calls? Was that the big idea?
03:35	Jason Lemkin	I just wanted to learn, like you guys did. I didn't think it would be as good as it was, because I used other products and other things, and they weren't as good as my digital AI was... it wasn't the...
03:47	Shaan Puri	Use case would be—what is it? What I said, or is...?
03:49	Jason Lemkin	The other interesting thing is the use cases with AI are much broader than you'd expect. I didn't expect people to upload their board decks before board meetings and ask for feedback—that's a niche use case. I didn't expect hundreds and hundreds of founders to use it as their "therapist" for their company when growth has slowed. I didn't expect that. I didn't expect so many folks to use it to review their sales scripts, their SDI [unclear—possibly "SDR"] scripts, when there are specific tools to do it, right? I didn't expect it to be so good at reviewing VC pitch decks. It's so good at it—better than anything I've seen.
04:25	Sam Parr	"Could you get tactical for one second?"
04:27	Jason Lemkin	So you.
04:27	Sam Parr	Go to Delphi.ai. You uploaded the Sassers RSS feed, and it did all of the work.
04:35	Jason Lemkin	YouTube, RSS, Twitter, blog, blog... but...
04:40	Sam Parr	"But could I go to a health guru or a business guru who I subscribe to and love?" </FormattedResponse>
04:45	Jason Lemkin	Yes.
04:45	Sam Parr	Can I say "youtube.com/alexremozzi", upload that, and it automatically does it? Is this... well?
04:51	Jason Lemkin	That's an interesting question. It is prohibited, right? It is prohibited by their website—their terms of use forbid it. But honestly, this is an AI risk. You could do it. Absolutely—you could redo what I did. It might take you a little while longer, but you could just clone me exactly, and then you could do it just as well as I did.
05:09	Shaan Puri	Sam, I've never felt so used and abused. When Jess Ma came on the podcast, she said, "Yeah, actually, I have my team training an AI of you, so that if I'm brainstorming ideas I don't have to ever call you. I'll just ask the AI." She continued, "Basically, you'll just be in the meeting with us." Then she added, "We just took all the podcasts and we're training it on how you think—the questions you would ask, what you would say, reactions you would give—and now you're just there in the meeting." So she basically hired me without ever paying me, right? I have this corpus of content out there. I've actually put my brain out there in the public.
05:41	Sam Parr	But how do I do that? How can I just upload a YouTube channel and say, "this YouTube creator has great advice," and upload 800 videos so they will all be transcribed? I could go to Delphi.
05:51	Jason Lemkin	It may take a day or two to fully ingest all that content, but it will do it just from the URL of your YouTube [video].
05:58	Sam Parr	Just on Delphi.ai, so I...
05:59	Jason Lemkin	Could I—could I make a mini-shot? I think it is a great product. There are other people that can do it too. That's not... I'll tell you what I think is interesting about Delphi. We're looking at all this Cursor and Windsurf stuff and Lovable, and you don't realize—they're just so far ahead in B2B and with regular people only because they're better people working on them. If we had the quality of people working at Windsurf working on CRM, or working on marketing tools or podcasting tools, your jaw would drop today. But this is why your jaw is going to drop over the next twenty-four months: mediocre engineers cannot... the engineers at OpenAI are so good. They're just so good. The average engineer working at a struggling CRM, right, is just not that good. But they'll catch up. Yeah, you can build all of this. And here's the interesting learning: we could build a digital Sam, and we could do this by next week or whatever. You can do it. My learning from AI is: it's okay that it's not you. This is why, for example, you can train your voice—it uses Level 11—is that a product? Or Lovo? I get a little confused.
07:02	Shaan Puri	"Eleven, right?"
07:03	Jason Lemkin	Yeah — 11:11. Yeah, 11:11. And Level Labs is epic, right? But you can choose: do you want it to be "exactly like you," which is what Brian Halligan did? I realized that was creepy. I don't want Digital Jason to be me. I want Digital Jason to be a prime version of me — 90% of the time much better than me, but distinct. I don't want people to be confused, like Brian Halligan. And so it's okay that it's not Sam. It could be a version of Sam that is better than Sam.
07:34	Sam Parr	Alright. A few episodes ago I talked about something and I got thousands of messages asking me to go deeper and to explain. That's what I'm about to do. I told you guys how I use ChatGPT as a "life coach" or a "thought partner." What I did was upload all types of information: my personal finances, my net worth, my goals, books I like, and issues going on in my personal life and businesses. I uploaded so much information that the output is a custom GPT I can ask questions about the issues I'm facing. For example: "How should I respond to this email?" or "What's the right decision, knowing my goals for the future?" I worked with HubSpot to put together a step‑by‑step process showing the audience the software I used and the information I fed into ChatGPT. They had ChatGPT walk me through everything, so it's super easy for you to use. As I said, I use this 10–20 times a day. It's literally changed my life. If you want that, it's free. There's a link below—just click it, enter your email, and we will send you everything you need to know to set this up in about 20 minutes. I'll show you how I use it 10–20 times a day. Alright, so check it out—the link is below in the description. Back to the episode.
08:49	Shaan Puri	So, okay. Let's walk through these top 10 learnings. So you build this AI; it has 35,000 conversations with people... [sentence trails off]
08:55	Jason Lemkin	Did it, you ask? "More like... almost."</FormattedResponse>
08:56	Shaan Puri	So now you have 50,000 conversations and learning. Number one, you said, "users will tell AI things that they won't tell humans." Yeah—what do you mean?
09:04	Jason Lemkin	Yeah. Well, first of all, before we can get there—one obvious thing: there's no way I could do 50,000 conversations, is there? I'm not even a people person. I could do one or two a day and I get tired. Just think about the delta between two a day and 50,000. But yeah, here's the thing: there are some things that are somewhat unique about me. First, the amount of content—Sean, like your math, right—the amount of books we're talking about here, the amount of content I've written just about B2B, just about business-building and business software. The other thing is, there are enough people—I'm not like a super influencer like you guys are—but there's enough people that know the digital me that they already know what to ask me. They've already watched me at an event, they've read the content, so they ask me things without thinking it's a prompt. They just ask things. They want to talk about their deepest fears. Like, "I've partnered with Sam on this podcast, but I'm not sure he's as committed as me. I see him doing other things—he's into this accounting software thing and this Hamptons thing. I'm worried he'll quit my podcast on me. What do I do?" My AI is really good at talking that through. And so people just think...
10:08	Shaan Puri	To ask.
10:09	Jason Lemkin	Me things.
10:10	Shaan Puri	How does he know all this, Sam? "Hey, Jason—the heart there, Jason?"
10:13	Jason Lemkin	You think it's funny, but you might ask me, "Hey, Jason?" </FormattedResponse>
10:15	Sam Parr	"Jason, how about you shut up?"
10:18	Jason Lemkin	Well, now... there are thousands of these types of conversations: - "My cofounder is not as committed." - "I don't know if my board's gonna fire me." - "I'm only growing 18%." - "I have 11 months of runway," right?
10:30	Shaan Puri	"Were you 'Zuckerberging' here? How'd you know what they were saying, or was there some privacy here? How do you know what's being said?"
10:35	Jason Lemkin	It is interesting. On this app, you can set how much privacy you want — you can anonymize everything. Well... no, because there are different use cases. If you're using it as a support tool — think about it: using it as a support tool, you don't want anonymity for support. How can you follow up with someone, right? Right. Then there's a therapy level where you would want true anonymity. And then there's, like, a sort of... everything's a slider in AI, in a sense, right?
10:56	Sam Parr	Dude, I think that—like you said—the therapy thing, as if it's "Can you believe that?" And yeah, I'm like, "Dude, I use ChatGPT for therapy three hours a day."
11:06	Jason Lemkin	Yeah, well, everyone does.
11:08	Sam Parr	"Like, the majority of the time I'm thinking, 'This email pissed me off — how do I reply in a graceful way?' Or, 'I argued with my wife, or I'm upset about that and I'm not doing this.' The other day, on Sunday, I saw someone who had a super fancy home and I was like, 'How do I get over envy? Do you know what I mean?' That's all I use it for."
11:28	Jason Lemkin	It is, and this is the same thing. It's just because it's got the 20 million words for business-to-business content, sales content, scaling revenue. It turns out this—and this was a surprise to me; I didn't believe it—it's better. Much better than ChatGPT. It's much better.
11:44	Shaan Puri	And you said something like, "you trained it afterwards." Can you—what do you mean by that? Well, how did you train it afterwards?
11:51	Jason Lemkin	Well, listen. I'm gonna embarrass myself, technically. Okay. But basically, those 20,000 words—you basically rag it. Okay, you've...
11:58	Sam Parr	20,000,000 (twenty million)
11:58	Shaan Puri	20,000,000 or twenty-five.
11:59	Jason Lemkin	I have 20 million words of my content. You turn it into a database and add that to ChatGPT, Claude, or DeepSync. The way you add it—and the relative weights of my content versus the generic content—radically changes the output. ChatGPT never asked my permission—it's already slurped up all my content, hasn't it? Every single YouTube video is in ChatGPT; everything I've written. I haven't been paid a nickel. What's interesting is that the content is already there, but by heavily weighting it toward my content and my type of knowledge, it's not just a little bit better than ChatGPT—it's much better. For example: how do I create a sales comp plan? I have four SDRs and five AEs [SDRs = Sales Development Representatives, AEs = Account Executives]. Our average deal size is $2,000 a year, and we want to go from $2,000,000 to $6,000,000 next year. My team won't really do outbound. ChatGPT is going to give you a very mediocre answer, but my AI is going to give you insanely good answers, even though overall we're pulling from the same content.
12:58	Shaan Puri	"And so you had to do that manually? Was that the training, or did you do more? Did you do more training than just that?" "Well, first..."
13:04	Jason Lemkin	Of all: it auto-updates every day, which is better than ChatGPT or Claude. It slurps up everything on my blog, on my social media, on my YouTube — every single day. It does it essentially almost in real time. Yeah, it does it itself. It auto-updates. Now, what I do do — and I'll tell you my learnings on my hallucinations — is I audit a few of them. I read some and see if they're wrong, especially the ones that are not anonymous. When I see something that's wrong, I fix it. There's a section where I just train it. I tell it, "No." For example, I said, "Sam and I first met in February 2016 after I wrote a Quora post, and he had me do a meetup at The Hustle," because it got that wrong — it made something up. I do that for about 10–15 minutes a day. It turns out it doesn't get very many things wrong. The things it does get wrong, it consistently gets wrong, and I'll tell you why. Then I just fix it, and the next day it's right. If you do nothing — you don't train it and you put no content in — this average VP of sales is gonna say, "It doesn't work." I logged into ChatGPT a year ago and it hallucinated, right? Like, "You're gonna lose your job."
14:14	Sam Parr	But the hard part here is... I am so committed to ChatGPT. It knows so much of my information because we've had so many interactions. Are we all going to have, like, files—our health records—where it's all in the background? It sounds like I get nervous to switch to any other model because I'm like, "Ugh, ChatGPT. I've already asked it so many questions; it already knows everything."
14:35	Jason Lemkin	It's an interesting question. Here's what I've learned. First, memory is a big deal. [“chatchippity” — likely ChatGPT] could already remember questions, but now it's remembering everything. So, it's a moat, from the top folks I've worked with on this. On the one hand, it's a moat—just like being in the Google ecosystem is a moat, or, back in the day, being in the Yahoo ecosystem was a moat. But the moat is not as deep as we think. Our ability to get trained and up to speed does not require 20 million words. So yes, it is a moat, but if you decided Claude was much better for you and you spent a month including your data each day, you're probably there. Maybe not 100%, but it's enough.
15:13	Shaan Puri	I had a good startup idea. I think it was Aaron Levie or somebody tweeting this. They were talking about "Plaid for AI tools." Basically, you'd be able to—just like the way Plaid lets you connect your bank safely to any financial service—to take your memory and context, what ChatGPT knows about you, and port that to some new app you're trying. You wouldn't hand over everything, but you'd allow it to pull from that. And I was like, "Oh, that's..."
15:38	Jason Lemkin	Ah — that's actually a very interesting idea: if apps will even exist in the future.
15:42	Shaan Puri	What will exist instead of apps?
15:44	Jason Lemkin	Why? We're going through a transition phase right now. We're chatting once again. We've been chatting since the ICQ days—since AOL. We've been chatting at our keyboards with people. Now, with AI, we're chatting again. But we're starting to not even know where we're chatting with. Right now, already—MCP is new—but today OpenAI and Claude can pull from other apps like HubSpot, Notion, and others, and just pull the data out. I don't have to go to Notion or even Google Calendar or HubSpot anymore, right?
16:13	Shaan Puri	Right.
16:13	Jason Lemkin	As we live in this... and then, you know, people think this, Johnny: I think $6 billion is crazy — it's a great bet. We're going to go from 20–22 minutes a day in ChatGPT to 24 hours a day in ChatGPT. When we're in ChatGPT 24/7, we're not going to pull up a CRM. We're not going to use the interfaces — these interfaces are already dying. You can see it, Sam, in the amount of time you spend in GPT. Then you go to a B2B app and you're pulling your hair out — it's so dated, isn't it? Yeah, it's so dated already. It's only 2025 and it's already dated to go to any B2B app. They're all going to die. How?
16:49	Sam Parr	Does the excitement in the rate of change right now compare to... I don't remember if you were involved in '99 or in the year 2000, but you're close, right?</FormattedResponse>
17:00	Jason Lemkin	Yeah, I was a kid. It was my first job back then.
17:02	Sam Parr	"Yes. So—because everyone talks about the dot-com boom, and I never got to... I mean, I never experienced it. But how is this today compared to then?"
17:12	Jason Lemkin	Those days were the first time — until AI — when people felt they could just do something out of nothing. It was miraculous. It was miraculous. And that is what AI is like today. People are just doing miraculous things, and it's exciting. But, man, it's super scary for incumbents. It's super scary for incumbents.
17:27	Shaan Puri	Do you think that everybody who is maybe a content creator—at least, yeah—is going to have this sort of, like, stunt double? This intellectual stunt double the way that you created here. Do you see, once you've done this experiment, are you like, "This is the future," or "This was novel but no, I don't think it's going to last"? Where did you land after the experiment?
17:47	Jason Lemkin	Well, first of all, what I think is: I'm not a consumer content creator. I don't think anyone is going to create content without AI going forward, and the tools are going to change. Like, for example, just for our last big disaster event, I did all the speaker promos in Higgs Field. Have I ever used Higgs Field? No — it's...
18:04	Sam Parr	"Incredible—how do you spell that?"
18:06	Jason Lemkin	Higgsfield.ai — it's the ex-head of AI from Snap. It's one of the coolest apps on planet Earth. Everyone's talking about this new Google video thing, right? But they were already doing it. All you have to do with Higgsfield.ai is: I can take a screenshot of us right now, put it in Higgsfield, and it will create a movie out of it. There are a lot of different ways, but a really simple way is to take an initial photo and an end photo and it will just connect them with a narrative. For example, I would take a picture of Yamini — like her HubSpot headshot — and then take a picture of a stage at SaaStr, and it would make a video of Yamini waving to the crowd and running on stage at SaaStr, being excited. A year ago, all the digital promos were static — you know, the classic static headshot with some chrome around it. Today, everyone got a video, a really cool video. Next year it will be me and Yamini talking for ten minutes about AI and neither of us will have had the conversation. It doesn't mean I won't exist, but of course that's already coming. We could almost do that already. So I did all these things now. Instead of waiting for our designer two months to do a static thing, upload it to Dropbox, and then I forget where it was and lose the URL and have to do it again, I just did it myself in Higgsfield in like ten seconds — so I'm... </FormattedResponse>
19:17	Sam Parr	Looking at Hicksfield... This is—this is insane. </FormattedResponse>
19:20	Jason Lemkin	It's beyond. It's so good—so good. This is the... this is why AI is exciting: the really badass people can do crazy stuff now. It used to take a badass person like six months to do a prototype and then two years to get it good. Now someone epic, like the ex-head of AI at Snap, can—what you can do at Hicksfield will blow your mind, right?
19:39	Shaan Puri	Dude, I could even tell just by the way you're talking. I feel like if we did the same podcast, you know, five years ago, your rate of talking would have been like 10 beats per minute lower.
19:46	Jason Lemkin	Well, dude, you're... there's, like...
19:48	Shaan Puri	A baseline level of excitement that has just risen—like our new blood pressure is here.
19:53	Sam Parr	"Well, as long as I've known Jason, you were borderline grumpy — you were pissed. You were pissed off all the time." Me: "Like, am I?" "Yeah, yeah, yeah. You were like... I remember you. You weren't tired, but you were tired of people asking you dumb shit. You always had this energy where you were talking to someone and you're like, 'You're making a mistake. Stop this. You're making a mistake.' You had this almost anger in you. Now it's the exact opposite. Now you're... you're optimistic."
20:20	Jason Lemkin	"I am... I am. You're right. I'd like to think of it differently, but you're probably right. The way you perceive me is important. It is. I am very optimistic about it, but the difference is I used to think all these products were durable—that they were durable. Now I don't think any of them are durable."
20:37	Sam Parr	So, what's that mean?
20:38	Jason Lemkin	When I first started doing this SaaS for content, I made up this expression, which I stole from the mayor of Tel Aviv, and people have used it ever since. > "0 to 1,000,000 in revenue: impossible — we don't need another product. > 1 to 10,000,000 in revenue: unlikely — there's so much competition. > 10 to 100,000,000: inevitable." From impossible to inevitable. We had a book that sold a couple hundred thousand copies; that was the theme. So you go through this stage because what I learned as a B2B founder.
21:01	Sam Parr	So you're saying getting from zero to a million is really, really hard. Yeah — getting from one to ten feels impossible, but once you...
21:08	Jason Lemkin	Get to "unlikely."
21:09	Sam Parr	Yeah, unlikely. But once you get to 10, it's inevitable to get to 100.
21:13	Jason Lemkin	If you had good founders, it was a punch card. Now it's disrupted. Now zero to one is very likely in AI. You build an incredible tool that no one else can do, and you can do a million in a week or two. But now everyone beyond $100 million is getting disrupted. There are so many software companies at $100–$300 million in revenue that aren't growing anymore. It never used to happen four or five years ago. You hit your punch card: you had 120% revenue retention and you had a brand. All you needed was a brand, revenue retention, and products changing about every five years.
21:46	Shaan Puri	Right.
21:46	Jason Lemkin	If the product changed every five years and you had a brand and tens of millions in revenue, that meant you had product–market fit. High revenue — like high net revenue retention (NRR) — meant it kept going as an engine. This 120–130% revenue retention just kept it going. It took you from $20M to $30M to $50M to $100M, and that’s why, in 2021, the average public SaaS company traded at 70x ARR. Public valuations at 70x ARR in 2021 are almost unimaginable today when they're around 5x. It’s because it worked then. Now, companies like Okta (growing 10%) and Salesforce (growing 7%) — they’re good companies, but it’s not as durable as it once was.
22:30	Sam Parr	So, are those companies my premise, or what I thought? I thought, because Morgan Stanley had this report—who gives a shit about Morgan Stanley—but they said that "the majority of value creation in terms of market cap for AI is going to be the big companies using AI to get better, versus new AI companies." Do you disagree with that?
22:53	Jason Lemkin	I think that right now, as we record this, there's a hint of truth to it — but Salesforce is not accelerating. Okta is not accelerating. Lots of folks are not accelerating. ServiceNow has a lot of AI, but it already had workflows that were automation. People point to that and there's qualitative proof of change, but we haven't seen it in the numbers yet. So there's a window for the big guys to leverage this, because the interface we talked about is dying. People are not going to want to use these dated interfaces anymore. Even if it's benefiting the big companies today, what is the value of a CRM if I don't log into the app anymore? If I just talk to ChatGPT and say, "Give me the 10 deals I have to work on this week," how valuable is one CRM over another? Will I even know I have a CRM, or will it just become plumbing — a database? One problem with AI: there's been too much nerdy nomenclature over the last year — too many O4s, minis, O3 maxis, LLMs.
24:04	Sam Parr	Dude, it's so complicated. It's like knowing the difference between an iPhone 13 and an iPhone 14.
24:09	Jason Lemkin	It's all going away this year. All that complexity made sense. As crazy as these numbers are, Anthropic went from $1 billion to $3 billion in revenue in five months. They just announced that growth, and that is basically all infrastructure — plumbing — revenue. In the area of infrastructure, you need to know the difference between the models and R1. We're just starting the application age. In the application age, we won't care what models HubSpot uses. HubSpot already has a lot of AI — do you know what models it uses? You don't. Right? You don't. So these nerdy terms are going to fall away. Think of MCP as the next-generation API. What MCP does is it lets any AI (or other systems) talk to applications without having to write inside the AI. Right now it's just rolling out. If you set it up properly — it's still a little nerdy — from ChatGPT you can change your Google Calendar and you can work with Notion.
25:07	Sam Parr	"Tell me how."
25:08	Jason Lemkin	You connect it with me. You just—well, I can't quite do it with you online. Maybe in two months. It's literally when you OAuth in: it will do it automatically. But I can show you offline. You just connect it to your Google Calendar and then... there, there, there. It's still a little bit of a headache around authentication and security. There's a little bit of magic to go, but not much. Then you can just talk to your calendar and chat. HubSpot has announced this, so Dharmesh...
25:33	Sam Parr	"But what are you using to connect the two?"
25:36	Shaan Puri	MCP is a protocol, right? So it's like a language. Imagine this: we have English, we have Spanish. Then, when the internet browser came out, it was like, "Hey, there's a new language — we're gonna call it HTML," and you had to learn that language if you wanted to tell the browser what to do: how to show a web page. Now MCP is a new language that came out that basically says, "Hey, if you're one AI app and you want to talk to another application, how are they supposed to talk to each other?" MCP became a protocol that any app can use. For example, if HubSpot wants all the AI tools to integrate with HubSpot, they don't want to do a one-by-one deal and build a technical integration with each one. So they say, "Oh, cool — we'll also speak MCP. We understand that language; we know how to do that." It's basically like how APIs work for most websites, but MCP is the AI-specific version.
26:25	Jason Lemkin	Aside from that, right now to integrate with—and I haven't ever used HubSpot's API, but I'm sure it's a great one—that is bespoke. Each one is different. HubSpot's API is different from Salesforce, Notion, and Linear. You have to learn it. You have to get a key. You have to code to it. You have to understand its nuances. It's a lot of work, right? Like, you may know, startups will be like, "Why don't you have a HubSpot integration?" Well, because it's work.
26:48	Shaan Puri	Right.
26:48	Jason Lemkin	With MCP, it's early, but it's just starting. It won't be any work—you just tell MCP to go talk to HubSpot or go talk to Notion, and it will just do it. All that work goes away. So if your AI can talk to any app you use, take all the data out and let you talk as a human—for 99% of the world, we won't use HubSpot the way we use it today. There's no reason to log in to this app to figure it out: how does the UI work? What are these tabs? How do I configure a tab? And the kids these days will never know this. This generation coming up—they will never use software like we use it. They will never use it.
27:25	Sam Parr	What is the office of OpenAI like right now? You were saying, "well, it sucks right now, but in 60 days it's going to be the best," which is insane. So, how many people work at OpenAI right now? And actually — Sean, for you as well — have you guys been to the office? What's it like there? What's it like at that office right now if they're able to do all these things so fast?
27:46	Jason Lemkin	I think it looks like they have about 4,500 employees. I'll tell you what's super interesting: they only have about 60% employee retention over two years.
27:54	Sam Parr	Because everyone’s going and doing their own thing.
27:56	Jason Lemkin	Probably. But it's just... if OpenAI only has 60% team retention, what's your hope to keep your AI team? You'd better treat them really well.
28:07	Shaan Puri	So I have a friend who told me about the early days of Twitter. He had built a company scraping Twitter and thought, based on what people are saying, I could figure out what's actually trending in the world. > He explained his method: people always talk about New York, so if I hear "New York" a thousand times, that doesn't mean New York is trending — it just means New York is already big. But if suddenly everybody's talking about Tallahassee — which normally doesn't get talked about that much — and it's being talked about as much as New York, that means it's trending. > He had a separate site that tried to figure out what the trends on Twitter were. They ended up getting bought by Twitter, and that's now what you see when you go on Twitter: "Trending." I asked my friend Abder what it was like in those days. He said, "I thought I knew what a startup was." He related to me so much: he thought a startup meant waking up every day and trying to get growth — pushing, pushing, pushing the product and marketing out into the market. Then he gave a vivid image: imagine you're pushing a boulder up a mountain and you look up and there's a giant avalanche coming straight toward you. For his first six months at Twitter, he said he woke up with the QWERTY keyboard imprinted on his forehead — he'd fall asleep at his desk and wake up back at work. That's how he lived for six straight months. He said everyone was making fun of them because Twitter had the "fail whale," but nobody had ever scaled a service that fast. They couldn't keep it up. People thought they were lazy or stupid, but he insisted, "We're the smartest people — we're trying our absolute hardest — and it was just so, so difficult." I think what's happening with a lot of these AI companies is that, you know, Sam, you and I — I think we probably only really experienced the "push" phase for most of our startup careers.
29:58	Sam Parr	Yeah, it never felt like I was holding on. It felt like I was pushing, and it got a little easier sometimes, but it was always resisting.
30:05	Shaan Puri	Like, I have this company now — we haven't announced this — and we have this AI bot that will come and tell us our new contracts and our new revenue. Last week we closed $1 million in new contracts in a week. I have never felt this level of pull in any business I've ever done. We're not working 10 times harder; it's just that the market wants this type of product at a different level. I'm sure that folks like, you know, OpenAI and all these AI startups are experiencing a startup experience that was not anything like what you and I experienced, Sam.
30:39	Sam Parr	I think it was 0 to 800 million users in seven months. Is that what Mary Meeker said?
30:45	Jason Lemkin	"Yeah, I think so. I just wrote it up on Twitter. We could look on X, but yeah."
30:49	Sam Parr	Ten percent of the world. Ten percent of the world in eight months, or something like that.
30:54	Jason Lemkin	That's why the rate of change is so high. Everyone who's a luddite or a "Debbie Downer" is missing the...
31:00	Shaan Puri	Right. Before, it looked like it was a bunch of looky-loos—people were going and testing the product, but it wasn't really high retention. Then they've shown that, in fact, it's not just that the number of users has doubled; usage per user has also doubled. So you have a double of a double. It's pretty obvious to me now that I'm kind of an idiot for not investing in this company, even at really any price. This is basically the Facebook of this generation, right? Yeah.
31:29	Jason Lemkin	I don't think we realized that OpenAI's ChatGPT would have an 85% market share at the prosumer/consumer level. That would have been a surprise if we had known. Ironically, ChatGPT was an experiment — a proof of concept. It wasn't really meant to be the product or the center of OpenAI.
31:46	Shaan Puri	"Like, do you beat yourself about this too, the way I do? I just look back and I'm like, how many hours have I spent thinking about AI? I'm supposed to be an investor, you know — like, you're a SaaS investor. This is the greatest SaaS, you know, the fastest-growing SaaS business ever. Are you an investor in this?"
32:00	Sam Parr	"And did any of you have the opportunity to invest in it?"
32:03	Shaan Puri	I did, but you can hunt it down, you know. That's on... that's—
32:05	Sam Parr	Oh, good.
32:05	Jason Lemkin	Yeah, for me—I want to, life is short. I want to own 10% or more of a startup and have fun. So hunting something down just is not my vibe, but it's the way to make the most money. Don't get me wrong—hunting it down and [unclear: "shot or my first million"] and just getting some money in, like, that's the smart way to do it, but I...
32:25	Sam Parr	Bet you — I bet you, Dharmesh. So, I think on the podcast Dharmesh... I don't know if he explicitly said $15 million, but he sort of winked at it when we mentioned it. He owned Chat.com, which he sold to OpenAI for what he paid for it — which I think was around $15 million — and he got OpenAI's ChatGPT stock. Yeah, there's a world where that becomes worth, like, in the ballpark of his HubSpot stock, I think.
32:48	Jason Lemkin	You know... that does probably, right? I—it's none of my business, but I remember thinking about Aaron Levie, who's one of my favorites. He was early in Stripe and early in Gusto. Back in the day, Aaron had a window before Box went public where everyone went to him and he had a little time. He just did all of them. I'm like, "Man, if this guy did Stripe in the seed and got all that... how—" That's gotta be none of my business, but it's gotta be worth even more than his Box stock, doesn't it? For sure.
33:16	Shaan Puri	He lost 4% of Box, I think, when it went public, and he's been grinding Box for like 20 years.
33:21	Jason Lemkin	I love him for it.
33:22	Sam Parr	Yeah, wait—so he was just throwing around the money on the Stripe investment.</FormattedResponse>
33:26	Jason Lemkin	There was a window. Listen—it's none of my business, but there was a window. I've been doing this long enough, right, where before Box IPO-ed he had time. Then, once you're public, you just can't do this stuff anymore. Maybe Marc Benioff can. And so, everyone—he was the guru, right? He was the one guy. Everyone wanted Aaron [Aaron Levie]. So he would write these checks... you know? I know—we could look it up—but I know he invested in Stripe and Gusto in the seed round, and a bunch of others. It's none of my business, but I'm like, "there are these moments in time that could be worth more than all of his Box shares." That just gives me even more respect to keep going, because so many founders now have a different relationship to money and investment than they did a few years back. A lot of founders would quit today if their investments are worth more than their founder stock. A lot of them do quit. A lot of them do quit now.
34:12	Sam Parr	You know what this reminds me of? At The Hustle, I think it took me three years to get to a million subscribers — a million users. That was life-changing for me. To see how fast some of these guys are growing... it's literally like that per hour. It's sort of like when you see a billionaire who's worth about $4,200,000,000 — and you're like, "Man, that... two divided by 10 would be awesome for me." Do you know what I mean? It's hard to grasp how big these things are and how fast they're getting that big.
34:50	Jason Lemkin	Well, look. It's slightly under-discussed, and it's a crass topic. You might have discussed it, Sam, otherwise... but, like...
34:55	Sam Parr	It's a crass topic.
34:56	Shaan Puri	"Sounds like it's right up front."
34:57	Jason Lemkin	But because OpenAI will be worth...
35:00	Sam Parr	"You're a filthy pig."
35:01	Jason Lemkin	No — but when HubSpot... I've been around long enough to remember when HubSpot IPO'd. It IPO'd at about $800 million valuation. Today it's worth like $40 billion. But now, like, OpenAI will be worth $1 trillion. Is that — the rich are so much richer, right? The regular person is a little bit richer in tech, and that's great. But the big wins are 100 or 1,000 times bigger than just ten years ago. Just ten years ago, the number of billionaires in SaaS and cloud was much smaller. I wrote up on SaaStr the number of billionaires in SaaS and cloud — I could post it, but your jaw will drop how many billionaires are in it, just because the markets and numbers are so much bigger. Right now there's already about 100 billionaires in B2B software. I wrote up 100 billionaires.
35:43	Sam Parr	"For the people listening: what personality types—who's poised? What does the 22-year-old today look like who can pounce on this? What's the profile or personality type of these people who are just excelling like crazy?"
36:05	Jason Lemkin	I think it's just two things, and it really probably hasn't changed since Bill Gates' days starting Microsoft. First, you have to be able to ship insanely good software. Insanely good — like high-fidelity software that makes you feel you're using ChatGPT. This is not trivial stuff. Second, you have to be relentless about owning a market. There was a while in 2020–2021 where being third or fourth was great. Things were so good that you'd think, "Well, if I'm number four in the market, I don't have to sell a lot of stock, it's calmer, and I could sell my company for $400,000,000 — that's better than IPO-ing." Now, the best founders are relentless about being absolutely number one: destroying and owning the market. Because markets are bigger, that position compounds into something crazy. As one example: Sammy and I are both investors and owners. Adam — whatever his background, even a high school dropout starting this company — is going to disrupt the entire restaurant industry. He's already on the way there. He won't be number four. He has an incredible technical cofounder and an incredible engineering team. Here's an interesting thing: I invested in the seed round [2021]. "Adam hates it when I say this," but the product was not very good. Dean — the CTO — did a lot for the product, but it was not very good initially. The reason it crushed all its direct competitors (not the big indirect ones yet) is that the software is just so much better. That advantage compounds every quarter, and AI is accelerating that even faster. So: you have to build epic software. If you destroy markets, that's the path. I just don't think this is a good time to be number four. I could be wrong, but that era had its time.
37:54	Sam Parr	I actually disagree with you. I... I think I understand your sentiment, which I agree with, but I disagree with how you're phrasing it—that it's not great to be fourth or fifth place or whatever. We've had a bunch of people on this podcast who built apps that were really fast money grabs; they got really big really quickly with tiny teams. What I would tell them—who the hell am I to say?—is: "This doesn't seem durable." So keep your team small, ride the wave, take all the money, and do something else eventually. It's a little bit different than it used to be because you don't need to build these huge infrastructures. You can have one-hit wonders that crush it.
38:30	Shaan Puri	"One-hit wonders" is the wrong phrasing, but I know what you mean. A base hit now can (a) provide you enough cash flow to change your life, and (b) set you up with five more doors—whether it's investing or reinvesting into another company you start. You don't have to build a 20-year durable company in order to be a winner anymore. I think I understand what you're saying. Back when we were in San Francisco early on, the small lifestyle business would get you the equivalent of a job. You would make a couple hundred thousand dollars ($200,000) if you had a lifestyle business — it was working, but it was small, solo, and bootstrapped. Now those same solo bootstrappers have about $8,000,000 ARR. It's just a different scale. </FormattedResponse>
39:15	Sam Parr	I remember. That's what...
39:15	Jason Lemkin	You know, the difference, though, is that... I think that term is something you have to be thoughtful about. In the age of AI, when there's so much competition, you can build a two- or three-person company that's small. But lifestyle businesses are being slaughtered. When these three kids come to SF, working in the Mission, and they're working eight days a week, and they've taken your little idea but made it much better, your lifestyle business is going to be unemployment. So this term, I think, needs to die. It made some sense a few years back, but I'm worried people are going to take the wrong lessons from it. You started this conversation: you can do more with fewer people, but you better work harder if there's competition—or find a space with no competition, right? But the weird thing about AI is there's competition in spaces that, two years ago, had very little competition.
40:01	Shaan Puri	Like, what's an example that comes to mind?
40:03	Jason Lemkin	I'll give you an example. When I started investing, I did a couple of investments in legal. I knew a little bit about legal, okay, and no one wanted to invest in legal. They were like, "Oh, it takes forever—lawyers don't buy anything; it takes forever. It's boring. It's a slow market." Now there are about 500 AI legal startups, and there's reasons for that, but you could not be at that sleepy pace. Same thing: some of the best investments I've done have been in support and post-sales. No one wanted to do this type of startup—boring: resolve tickets, pick up the phone. Right now there are thousands of voice startups; there are thousands of these. So if you think you can run a lifestyle customer-support startup, good luck. Good, good luck, because Delphi out of the box is probably better than you.
40:48	Sam Parr	You own a media company. You own, basically, a trade show that's incredibly successful. So, let's just categorize that as a "boring company" — one that is not a technology company. </FormattedResponse>
40:58	Jason Lemkin	For sure.
40:59	Sam Parr	Which is what I do as well, and Sean has done something like that in the past as well, I believe, for the people like that. "How are you using AI in a non-tech company? Are your employees now becoming significantly more efficient?"
41:13	Jason Lemkin	I mean, we only have two people now. I take it back — we have five people, and we do 5 million per person now. Did you...?
41:19	Shaan Puri	Have more people before that.
41:20	Jason Lemkin	We had much more people before. We had people in 2020 running — I didn't even... We had four designers. Now we have a... a little bit of a designer in AI tools. We used to have five people on this content team reviewing sessions; now we have zero. Now we just have AI. So we have none of the designers, none of the content people. All the ghostwriters are gone. They're the worst, right? Well, I write my own content... let... </FormattedResponse>
41:44	Shaan Puri	"Me: 'Drop you off at the airport?'" "No, but what?"
41:46	Jason Lemkin	Nine months ago, Yamini would come to a Saster event—she's done it three times—and she would speak. Three months later, a ghostwriter would write up a terrible summary of her talk. It was just terrible, and I would cry. We'd ask them to fix it because I can't do everything, right? About nine months ago, they got better. They weren't great, but they got better. I hadn't actually used Claude, and then I went into Claude. I'm like, "She just put it in Claude." They started charging us $5,000 a month for this. So we can do this for $20 in an hour instead of $5,000 and waiting two weeks with their clients, right? So we didn't need the ghostwriters—we got rid of [them] first. Then we had five people, then we had an agency we worked with for years. For a while it was great: they would review all these speakers. There's a lot of work to work with speakers—Sam, you've done it—and when you have hundreds and hundreds of speakers, it's impossible to do it all right. But they decided they just didn't want to work that hard anymore. They wanted to charge us two to three times more and only do half the work. So it just didn't help us, right? So then we're like, "We're SOL now." We're five months before Saster Annual and we just said, "Well, let's have our AI do it." It reviewed 300 sessions, 300 slide decks, 300 presentations, but do
42:51	Sam Parr	You have to chase those 300 people down to give you the...</FormattedResponse>
42:55	Jason Lemkin	Yeah, we still have to... It doesn't get rid of everything, but it does 90% of the work, and it's better. It does 90% of the work three times better. 90% of the work three times better.
43:04	Sam Parr	Was that you—the business owner who had to architect and come up with all these solutions?
43:08	Jason Lemkin	Between Emilia, who runs Saster — this media business and events — and me, we came up [with it]. No one else was motivated to do it.
43:15	Sam Parr	Of course.
43:15	Jason Lemkin	That makes sense. And then, we have a little bit of a sales team, but our AI just started to do all the screening and initial sales conversations for us. That helped a lot.
43:28	Sam Parr	Even for—this is a very niche business. Yeah, trade show business. Exactly. To say you do $25 million in revenue with five employees is pretty breathtaking, because I would have thought that would be on the lower end. Your business would be on the lower end of, like, "well, I still need people," but you have just proven that not to be true.
43:47	Jason Lemkin	It could do better with more people. It could do better with more people, not off.
43:52	Shaan Puri	Your life gets worse. No.
43:54	Jason Lemkin	To be honest, I don't mean to get off target. The problem is I just can't... Sam, this is why you think I'm grouchy because of an earlier conversation we had. I've just struggled to find enough A‑tier talent that wants to do this non‑sexy stuff. I would love to have 20 people on the team tomorrow. I have the budget for it—or at least 15. But I don't need someone who shows up to a meeting with a sponsor and doesn't know what we do. I don't need a designer who doesn't finish the design until after the event. I had that once in the past: they sent us a big bill but didn't finish the assets until after the event. I just don't need that. I don't need someone managing 10,000 people at our SaaStr Annual [SaaStr Annual conference] who forgets to arrange catering. I'd rather have the AI come up with the catering—the AI did our catering schedule this year.
44:41	Shaan Puri	Could you talk about the start of SaaStr real quick? When you started this thing, you had sold your company, right? Was it after you sold EchoSign, or a...?
44:50	Jason Lemkin	A long time ago—2012. But yes.
44:52	Shaan Puri	So, in 2012 you sell your company — great exit. As I understand it, you're like, "I've retired twice; I got in shape; I picked up a hobby," and realized, you know, like, you still love building. So, you sell your company — do you retire right away? Basically, do you take a break? Well... after...
45:09	Jason Lemkin	There's probably a little bit of parallel to Sam here. Both times after I sold my startups, I sort of didn't work for the better part of a year and just got restless for different reasons. I just got restless, and in fact both times after about a hundred days—hundred twenty days—I got a little depressed. Maybe "depressed" is too strong a word: lost. Lost, right? Definitely lost. I remember, you know, my second startup—Emergence Capital, which is a very successful B2B venture firm—had a CEO meetup every year where they would come. In my class, my group, I had David Sachs and Aaron Levy and Peter Gaster from Viva and Renee Assert. After I got acquired, I got to go to one more, and they said, "Can't come back anymore; you're not... you're not part of the CEOs anymore." I was off the team. You know, Ben Chestnut did an interview just a couple weeks ago with Kleiner Perkins and said the biggest issue when he sold Mailchimp for $12,000,000,000—or whatever it was—was he knew he would be instantly irrelevant. He knew he would be instantly irrelevant. He said, "I am now—I'm irrelevant now; Mailchimp's of the past." He's like, "You wouldn't even build a tool like Mailchimp today in the age of AI," right? Maybe that wasn't quite the...
46:18	Jason Lemkin	So yeah, that was it. But what was fun for me — and this is a long time ago — was that I was the first of those CEOs, like the Aarons and the David Sacks, that had an exit. I didn't have to pretend anymore. So I... maybe Sam thought I was grumpy, but I just shared every mistake I made: how I screwed up my first VP of Sales, how I screwed up meeting customers in person, how I screwed up marketing. There wasn't much content back then. It's a little different than today — there wasn't much content, so everyone just started to read this.
46:43	Shaan Puri	Were you being strategic? You're like, "Oh, I'm gonna start doing content marketing, and that's gonna lead to this, to lead to this," or you're just like, "I've got time on..."
46:50	Jason Lemkin	My hands... I've got stuff to say. I want to start blogging. I did—what? Now people give advice to, but I didn't; there was no one to give the advice to. I just wrote one blog post a day on a mistake I made. That was my paradigm: a mistake I made as a B2B founder that got to tens of millions. Back when Quora was a platform [Quora: a Q&A platform], and people listening to this won't even know what it is—Samuel—I answered one question. I answered one question. Peep this: it was a way for people to ask me questions, and those were great because they were very tactical and very specific. So I would do one a day, every day. It's a little different today with AI, but generally speaking, if you do that it compounds. It doesn't compound the same way revenue does, but if you do my "first million every week" and you just keep working at it, it does seem to compound. Not quite as linearly as we might hope, but it does compound. I didn't know—there was no advice back then. I just did one and one a day.
47:43	Sam Parr	"You're also one of the first people to do Twitter. You're one of the first tech people to—well, for a lot of tech people, Twitter didn't boom until COVID, but you were doing it very early. It was like you and Gary Vaynerchuk."
47:57	Shaan Puri	Twitter started with tech people. What do you mean?
47:59	Sam Parr	I don't think it was taken seriously. I think Jason looked at it as a craft, where he was like, "I'm gonna grow a huge audience here." I think Sean thought that, for a long time, Twitter was an afterthought compared to a bunch of other social media platforms. Now I think it's bumping.
48:12	Jason Lemkin	I just think I approach it as a "microblog," which is what Twitter in the old days called it. So I would put valuable content on Twitter instead of being grouchy or just... peep peep. I was early to putting valuable content on Twitter. If I benefited anything from content, I think it was just being early.
48:29	Sam Parr	You were early. You've been—dude, you are an output machine, and the stuff you write about is lyrical. You write about things—SaaS is a stupid, boring thing. You write about stuff where I'm like, "I don't even own a software company, but I wanna..." I love reading what he— you know? It's kind of like dye workwear with clothing. You don't have to know about clothing, but when he talks about ties, for some reason it's kind of exciting.
48:51	Jason Lemkin	Well, my actual, real goal in doing it was to make building boring business software more fun. Not in a jokey way, but to celebrate the fun—the exciting parts of it. That was my version of what Sam said.
49:04	Shaan Puri	Did you think, "Okay, then I'm going to be investing — this will help with deal flow"? And where did the conference come in? That's a pretty gangbusters business. I can't imagine you thought at the beginning, "We're going to be doing $5,000,000 per employee; we're going to do $25,000,000 a year in this business," where we have kind of a little monopoly — we're the big fish in this small pond of basically thought leadership for SaaS businesses.
49:28	Jason Lemkin	Well, the point is there are more thought leaders today. But—if you... I mean, it's a while ago. If you're curious, I just wrote this because I was bored, and it was something to do: something to share, something to add value to the world. No goals. No goals. I certainly wouldn't start off with a blog today if I were starting off today, right? As archaic as it is, there still is SEO, but...
49:50	Shaan Puri	What—what would you do?
49:52	Jason Lemkin	Even—it's not me, right? But of course I would do video; I do video. It's just not my natural medium. I'm a good writer. I'm not like Ernest Hemingway, but like Sam said, I have something special: I'm able to convey ideas in a way that helps people in writing. I still think today that's a rare skill, right?
50:09	Sam Parr	Dude, I hate video. Sean's pretty good at it. Sean can talk to a camera by himself and succeed. I hate it. I hate video. I think that... I think there's a generational gap.
50:18	Shaan Puri	"You need to put a supercut of Sam's ad reads."
50:22	Sam Parr	My God, they—Sam's suck. [unclear transcription: "they sam's" may be misheard]
50:23	Shaan Puri	Ad reads.
50:25	Sam Parr	It sucks, man. I'm... I'm incredibly uncomfortable doing it, I think. But I also think that the 25-year-old today—just because FaceTime has existed—they're so much better at it. There's a huge... there's a generational gap that I'm, like, four years too old to have caught.
50:41	Jason Lemkin	But if it's of interest, it might not be. The investing was an accident, too. It was an accident. I actually—for a little less than two years—worked at a VC firm that had invested in me. They recruited me to help them. I had no real interest in investing, and I showed up thinking, "I don't know what the hell—this is not me; I don't know what to do here." Luckily, I'd started a blog and founders started to come by the office. The founders of Pipedrive came by the week I started, and that was my first investment; it sold for 1,000,000,001.5 [transcription uncertain]. Then other founders started coming by. The founder of Talkdesk came by — that one was worth 10,000,000,000 in the last round. Then the founders of Algolia came by — that's worth 2.5 billion. Then the founder of Salesloft came by; we sold that for 2,300,000,000 cash in 2021. They kept coming by, and I thought, "Well, let's just invest in these ones because these are the best ideas I got." I'm not the outbound guy — "find Sam" — so I just invested in the five best guys that came by the office the first year. They all ended up being worth a billion dollars. So the weird thing about me for investing is I don't spend all my time doing it. I'm different than 99.9% of investors: I only do SaaStr super-fans who are high-intent inbound. I don't take favorited meetings. I don't do warm intros. If you tell me, "Sean, there's a startup that you want me to meet to invest in" or "Sam did," I would ask, "Do they desperately want me?" And they'd be like, "No, they haven't heard of SaaStr." I'm out. But if they really want me, they'll probably just email me. I don't even want a warm intro — I don't even want a damn warm intro.
52:14	Shaan Puri	That sounds like such a leak in your game, by the way. That sounds like an unnecessary bar, right? Not only do I need a warm intro where they say, "this person's kick-ass and they'd love you," but it's like—are they a huge fan? If they're not a huge fan, I'm out. That just...
52:28	Jason Lemkin	Seems... yeah. But the thing is: investing is so much harder than it looks. The odds that any seed startup is going to truly be worth billions of dollars are much lower than it appears in the media. It's really hard. If you're not incredibly intense—if you're not incredible, if you don't want this more than life—you are just going to sell your lifestyle business, which is very logical, right? </FormattedResponse>
52:49	Shaan Puri	So you're basically saying: if they're not really digging my content, they're not taking me as seriously as they should. Is that the implication? That if somebody is seriously trying to build a generational company, they should be loving our content and therefore they should want me on the cap table?
53:04	Jason Lemkin	I want them sending me a cold email that is so good that you would invest off the cold email. This is how I invest: I want a cold email that is so good — so good that you almost don't even need to meet them. The only purpose of the meeting is confirming that you want to invest. Like, every investment I've done that's good, I wanted to invest before the meeting. Every good outcome I've had — that's how I... </FormattedResponse>
53:25	Shaan Puri	I invested in "Owner" [company name unclear]. I'm a little bit after you guys, but I got an intro that was like, "this guy's one of the best founders I've ever met and ever invested in," and "this company is growing incredibly fast." Yeah — and then he sent a cold email basically right after that which said, "Hey, here's why I want you on board," and "here are my last three investor updates." In one of those updates, he shared that they had done a hack week. There were some Loom videos of the team sharing their hack week projects, and I actually just watched those.
53:52	Jason Lemkin	Yeah.
53:52	Shaan Puri	And it just shows the caliber of the team was really incredible. You could see the caliber of the team in a Hack Week project, right? Because almost by definition it's just one engineer and a designer. They come up with their own idea, have a very short timeline to ship it, and then have to pitch and present it. So you get to see the quality of the idea, the quality of speed of execution, and the quality of salesmanship all in one. Without abstracting away the founder, it's like that's the team itself. I emailed him back and said, "We don't need to meet — I'm in."
54:23	Jason Lemkin	Yeah — it's not the only way to invest, but what I can tell you is: if the email's mediocre, they're not gonna make it. There are exceptions. Okay, I know there are exceptions, but man...
54:35	Sam Parr	Wait, dude. So we're just at Hampton. We're hiring a bunch, and I'm putting together values to look for when hiring. I only made it—I think I'm only going to keep it to two or three values. The ability to write emails and communicate is one of the three.
54:53	Jason Lemkin	And I will tell you—one of the reasons we only have the five people is that I have lowered the bar at SaaStr in a way I never would as a software founder. I've lowered the bar again and again. I've hired folks where the email wasn't that great. I've hired folks who put an "e" in SaaStr. I've hired folks that didn't do the research before they started—100% failure rate, 100. Like, we already know this, right? But I've lowered the bar because I'd be like, "Well, this isn't cool enough; I gotta take who I can get," right? Always a negative—worth less than zero. Total, total zero.
55:25	Shaan Puri	Yeah, you have this great tweet. You said, "I think startups are about a great idea, but in the end they're about great recruiting." That's what you're talking about.
55:31	Jason Lemkin	They—yeah. And if you talk about "owner," not to use one, but it's because it's mutual. I will say there actually are lots of risks with owner, and I invested very early. You didn't see the two times we almost ran out of money. Now we have infinite capital, a lot of other things. But I actually do not know a more relentless recruiter than Adam. I don't know a more relentless recruiter than Adam, and that—that is a rare skill. We all learn it as theos; like, we get better. We always get better at recruiting, right?
55:58	Sam Parr	What does relentless look like? Like, what does that mean?
56:00	Jason Lemkin	Literally reach out to the 200 best people for the job. Talk to all of them. Set up meetings — cold, warm, lukewarm, direct via LinkedIn. Who are the 200? Like, talk to everyone in the world — who are the 200 best CMOs that I could possibly hire? Don't pretend to talk to 200. Don't actually only do five interviews and pretend you're doing them. Do 200. That's what S-tier recruiters do. I had an old co‑founder — I wish we'd gotten along better — but he was one of the best I ever knew. I'd roll into work every day at 8:00 AM, and he'd already handed me two candidates to talk to that were really good. Every day. I've never — Adam has that? — but I've never seen that magic, right? I don't have that skill. Still, every day he'd find a way: > "Hey, I've got this marketing/sales guy who wants you to meet. Can you talk to them today? Two more." But then they'd always be great. They won't always be a fit, right. And that's the only way you can scale as an executive. Unless you have a three‑person startup, you have a few...
56:55	Shaan Puri	More of these really great "how to be a great CEO" concepts. So, one is: to be a great — great — CEO, you have to enjoy telling the same stories again and again and again... hundreds of times.
57:05	Jason Lemkin	Yeah, yeah. Even Sam Altman does that, doesn't he? It's pretty much the same story, right? It is that one — the one I like that Sam does: "Comment — he tells you the future." Like, "Elon Musk, if you're listening."
57:19	Shaan Puri	What do...
57:19	Jason Lemkin	You mean by that... well, you know, he does this. You can make fun of this goofy video he did with Johnny Ive. I don't know if you saw it when they did it. Oh — perfectly lit. They're having coffee in North Beach.
57:29	Shaan Puri	My favorite rom-com of the year.
57:31	Jason Lemkin	"You think, 'Oh, this is silly,' right? But listen to what he's saying. He's telling you that, in the not-too-distant future, you will be on ChatGPT 24 hours a day. I'm making the bet that Johnny Ive is the guy to do it. I'm not betting that he's going to build a rabbit pendant. It's not that—I'm making a big bet that he knows UI and UX so well for the next generation that he will solve this problem of where we are. And Sam's already on ChatGPT more than the average today, right? The average person's on like 22 minutes. You said you're on it more, aren't you?"
58:00	Sam Parr	I'm on it all day.
58:02	Jason Lemkin	Yeah, so Johnny Ive is gonna figure out a way that it's in your ring, it's in your phone, and it's in your glasses — like Ray-Ban — and he's gonna figure out stuff we haven't figured out. It's in our... it's in Sean's and my headset, and we are just like, what if it was in this headset? Like, this is totally passive, right? I mean, I don't know if you guys have used Granola or the newest Notion. It records everything you do all the time in the background without permissions or visibility. You know why Granola is so successful? You buy it for note-taking versus... I don't know if the one you said does this, Sean, but Granola just records everything at the hardware level. Like, it could be recording us on Granola right now. We had a Chief Customer Officer Summit — 300 of the best chief customer officers at SaaStr this year — and the guy that put it on that day wrote up the whole summary of every single session that day. It was very, very good. I even— I couldn't do it. He's like, "Oh, just Granola — it listened to everything and it wrote it for me," and it was done that day.
58:52	Sam Parr	Do you feel this way, Sean? When I hear this, I'm like, this is so exciting. But the other side of me says, this is so intense. I have to... I hope I can invest in something and make a profit. Otherwise I need to take my winnings and go home and hopefully they compound by themselves. You're painting an intense picture. I don't know how I feel. Sometimes I'm like, this is so exciting — I want to get in; this is amazing. But other times I'm like, if my life depends on this, I don't want to play this game. This seems so intense.
59:21	Shaan Puri	I know exactly what you mean. I've now experienced basically... I graduated in 2010. When I graduated, that's when the light bulb turned on. I don't even remember life really before that. I didn't even know the Great Financial Crisis was happening—I was sitting at a Chick-fil-A somewhere in 2008. I was new to the whole world when I came out in 2010. Three things have happened since then that were of note. The first was mobile. Right when I moved to San Francisco, I joined Birch's startup studio. There was a single mobile developer. On my very first day he was giving a tutorial to the other twelve engineers—super seasoned Silicon Valley engineers—but they had never built an iOS app before. That ratio—one mobile developer to, you know, five, twelve, fifteen non-mobile developers—within twelve months flipped. It became: if you're not doing mobile, what are you even doing? Why would you do a startup if it's not a mobile startup? By the time you were a year into the job, it was a dumb idea not to be on mobile. I saw mobile come. Mobile was one of those shifts that was obviously going to happen, but it was a fog of war—you didn't know where the opportunities were. And actually, the opportunities were very different than the previous wave. So, like Michael, who had made a billion dollars in Web 1.0...
01:00:36	Shaan Puri	Or two. Or whatever. You know, the Facebook era—he didn't realize... even Facebook didn't realize. Zuck didn't build mobile apps, right? He was like, "No, no, it'll be mobile web responsive; that'll work." And, you know, the fact that basically Uber was the biggest winner during that era when we launched. Uber was a physical-world interaction: you push a button, but a car has to show up and you have to manage these drivers. That looked nothing like the winners before. So, we all knew mobile was big, but we didn't understand—the fog of war prevented us from seeing the path to victory.
01:01:09	Jason Lemkin	Yeah.
01:01:09	Shaan Puri	Then the next one was crypto. Crypto had a different flavor: it was not a consensus that this was going to be big. In fact, it was very fringe, and you looked a little crazy for saying it was going to be big. So most of us missed it for that reason. Now this is really the third one. I'll take COVID [the pandemic] out because that was a little bit of a different thing. This is the third one where, again, everybody agrees this is going to be huge — the biggest of all of them, the biggest opportunity ever. So everybody agrees on that, but there's complete fog of war. Like you said, even the apps that are working get disrupted. There's so much creative destruction and disruption. One moment you think it's the models that are going to get all the money, then you think it's NVIDIA that's going to get all the money, then you think it's the applications that are going to get all the money. Nobody knows how to win. Even on the investment side, VCs are in the same position. You're sitting there with this really weird feeling because you know this is the time when all the money's going to get made — the generational money's going to get made — and you still have no idea where to put it. People are trying to figure that out in the fog of war.
01:02:09	Sam Parr	"It's like that Warren Buffett phrase where—where he says, 'you see, he's swimming naked when the tide comes in or goes out.' The idea being when there's turmoil, that's when you get greedy. But the hard part is, a) the courage, and b) to know when the turmoil is actually happening. Are you in that? Now, we all know it's happening—this is the time. But do you or I have the want, desire, courage to actually get after it? Half the time I think, 'Hell yeah, this is amazing.' The other half of the time I think, 'This is scary.'"
01:02:41	Shaan Puri	I don't read a book.
01:02:42	Sam Parr	Yeah, yeah. I want nothing to do with this, you know. Do you guys agree?
01:02:45	Jason Lemkin	There was someone — I wish I knew who it was — someone on Twitter had a good tweet that resonated with me on this. He said: > "The smartest play, if you can possibly pull it off today, is to go make 5 or 10 million, cash out, and just chill, because you can't predict where it's gonna go. Whatever you can do to liquidate your assets now — they may be worthless down the road — so liquidate them right now and chill if you don't see the future, because this is the time." I do think that. I also think... you know, I used to — it wasn't until we ran our own AI — that my view changed. I used to think Vinod Khosla — I mean, obviously off the charts smart — but when he talked about how AI was going to bring mass unemployment in tech, I used to think he was being crazy. Now I can see it. It's coming. I can see it: McKinsey just laid off 10% of their team, but it's just a start because they put all of their learnings and all of their data into their own AI, like [saster] [unclear]. Now they just ask the AI; they don't need the kid anymore. Right? Half of our sales and marketing teams are going to be gone in two years. There will be no BDRs in a year. This mass change makes it hard to make bets. What's happening in venture, for what it's worth, is really two things right now. Seventy percent of the money is going into growth. Right now it's a great play: you close at a billion and then it goes to three billion, and the valuation goes five‑x. You put in $200,000,000; that $200,000,000 goes to $600,000,000. You take home 20% of the profits — you make $80,000,000 in a year. That's where venture's going: all growth. And then they're doing seed and early stuff, but implicitly people are assuming a very high loss rate and they're paying very high prices: $60,000,000, $100,000,000 valuations at seed, and they're hoping it's worth billions. They understand there will be uniquely high loss rates too. They're not saying it, but they know it.
01:04:27	Shaan Puri	Yeah, and that seems hard to make money, right? Those two—the combination of high valuations and high loss rate—doesn't that?
01:04:32	Jason Lemkin	Well, if you have $100 billion outcomes—if you get one per fund—that's the bet they're making: that there'll be a $100 billion outcome for each venture fund. Back when "hubs" got an IPO [unclear reference], the bet was you'd have a billion-dollar outcome per fund. One. Now it's $100 billion per fund. So if you believe you'll do that, you can do a little analysis or a waterfall, and you can just have Claude do it for you. You don't even have to do the analysis yourself anymore. I did this over the weekend. You just have to find one of them; the rest can crash and burn, right? So what is your game plan? I have decided that, whatever brand I still have—along with my excitement and my ability, and actually to know more because I've done my AI work—I actually know more than most people who aren't on the programming/tool side. I've decided that I'll get at least one $10 billion deal done in the next three years—two years, one, no matter. I'm only in it for one. Whether that means I have to make 20 investments or just one more, it doesn't matter. I admit it: I'm in it for one, right?
01:05:34	Shaan Puri	At the seed stage, at the early stage — yeah.
01:05:36	Jason Lemkin	"I want to own — I want to try to get one deal where I own 10%, or close, that's worth $10,000,000,000. I've sort of done it once on paper. I've done a couple of billion‑dollar ones, and I've missed a lot."
01:05:46	Sam Parr	What would that be worth to you if you own 10% at seed of a company that ends up at 10 billion (10,000,000,000)?
01:05:53	Jason Lemkin	"Well, multiply: if it's just 10 — if it's just only 10 billion, that's it. But multiply it by 10%, so that's a billion, right? And then you keep 20% or 25% of that."
01:06:03	Shaan Puri	But there's going to be so much dilution along the right.
01:06:05	Sam Parr	"That's why I'm asking: you don't account for dilution?"
01:06:07	Jason Lemkin	No, I'm simplifying my goal. You could own more—like, you'll own more—and then there'll be investments that are highly dilutive.
01:06:13	Sam Parr	So, at a $10 billion outcome, at a seed stage, you are looking to make $200 billion.
01:06:19	Jason Lemkin	"Yeah, that would help."
01:06:21	Sam Parr	Okay, that'd be cool.
01:06:22	Jason Lemkin	I think—listen—if you really want to know how venture [venture capital] works, there are two ways to simplify it. One, it's a great job. Two, it can be a mediocre job. Or you can try to make at least $100 million. I think anything else doesn't really exist—there's a gap. You're either in it for the fees and the job. At the beginning you think, "Wow, I'm a kid, I'm getting paid $250,000, and I just get to take meetings," and that seems great. Then you finally become a partner, you make a couple million bucks a year, which is much better than some other jobs. But there's a pretty big gap between that and real upside. Unless you're great, it's really hard to make profits in venture or carry. So the game is to make a fucking lot of it. If you're in a big winner—or in multiple big winners—just do the math, Sam. That's why most VCs make nothing beyond a good salary, but there are a few who have made billions. If you're in three, four, or five of those $10 billion outcomes (which is very hard), then you can make a billion bucks. The founders hopefully make much more; it is fair, right? So you're either playing that game, or it's performative, or it's really just for fees and salary. There really isn't anything in between.
01:07:29	Shaan Puri	But you also have the media and events business too, so you also have this cash flow. I assume that that's...
01:07:35	Jason Lemkin	A very profitable business. It wasn't before COVID, but it is now.
01:07:38	Shaan Puri	Yeah, so you have that either way.
01:07:41	Jason Lemkin	Yeah, I mean—I’ve made enough as a founder; I have a safety net. I mean, it’s a... but I don’t. I actually view it as a net negative for investing. But it is a passion. We do a good job, but yes.
01:07:53	Sam Parr	Last question: I know you have college-aged kids. </FormattedResponse>
01:07:55	Jason Lemkin	Freshman, sophomore
01:07:57	Sam Parr	Yeah. What are you advising they're gonna do when they're saying, "You know, what should I do for a job?" If they ask you that, what would you advise them to do now?
01:08:05	Jason Lemkin	First, I don't think a lot of these kids will ever have a real job. They will find ways on the internet to make money. It is interesting — there does appear to be a gender difference, which I don't like, between boys and girls. My son's a sophomore now. Ten percent of his class didn't even want to go to college at all. They all had this vibe: "I'm gonna go to Eastern Europe where it's cheap and I can just make money on the internet. I'm never gonna have a real job." I don't think my son really wants to ever have a real job. He's super smart — much smarter than me at math. He's deep in AI. He's deep into [Coursera?] and [windsurf?]; he pays for Coursera out of his own pocket. He's running assembly language now so he can "code to the metal" and do everything. But he'll never have a real job. Like he says, "Worst case I'll work for Airbnb," so he's never gonna have a normal job ever. My daughter is at Stanford. The vibe at Stanford — this is a shocker to me — is that they all want to go to grad school now as freshmen. They're architecting their classes and majors so they can go to grad school and defer work. When I was in college people cared about grad school; then for about fifteen years the vibe was, "Why would you go to grad school? Go to the internet." Now they all want to defer reality. I don't know that there'll be any jobs, and I don't know that people want to work. I honestly don't think — and I don't mean this negatively — that a lot of the current generation want to work. They want to be creators; they want to do other things. Whatever made the three of us want to work, I don't see it in the next generation. It's very rare.
01:09:41	Sam Parr	"The women want to be educated, and the men want to..."
01:09:45	Shaan Puri	"Move to Eastern Europe and make money."
01:09:46	Sam Parr	Out of.
01:09:46	Jason Lemkin	"United" — I'm not sure [unclear]. It's the women who want to be educated; I just know her group does. That may not be true for everyone. But the men not wanting to work at all — like coming out of high school and just saying, "no fear," like "no fear." I don't know what it was like when you... when I, during my life, I was worried I wouldn't have a job at different times, right? I was worried during college... I was always worried. [ending unclear]
01:10:04	Sam Parr	I was a senior in college. I remember thinking, "Man, I would kill to work at KPMG and make $45,000 a year." I remember thinking that would be a home run. For my senior yearbook quote, I'm like, "KPMG in Nashville, Tennessee — if I could get $50,000 from Deloitte, that's a home run."
01:10:22	Jason Lemkin	But it's so... so. I think—I don't know that they're going to work. And then I think a lot of the 2021 generation isn't going to work again either. I'll give you an example. Someone I knew a little bit... there's a lot. So I did this exercise: I just lowered the bar for hiring. I did this exercise; I know I shouldn't do it, but I wanted to try it anyway. I reached out to folks who had that circle on LinkedIn—what's the circle on LinkedIn? "Out of work"—what's it called, "looking for work"?
01:10:45	Shaan Puri	Yeah, open to work.
01:10:46	Jason Lemkin	I reached out to folks that I only knew a hint of. Maybe I had met them backstage at a hustle event or something. I met them in passing—enough that I had a micro-connection. They'd been out of work for six months, and I gave them jobs. All of them, in the end, basically said, "I want to make $200k or more, only to do meetings or manage teams." None of them were willing to do any work themselves. None of them—the people at the circle.
01:11:07	Sam Parr	The young people—the Gen Z folks.
01:11:09	Shaan Puri	Open to show their.</FormattedResponse>
01:11:11	Sam Parr	Their expectations for salary or for compensation are so high.
01:11:14	Jason Lemkin	Well, there's that too. I'm talking about folks who have already been out of the workforce for six months. But yeah, the expectation is they'd rather not work. That's the thing: a lot of people would rather not work than get their comp [compensation] right.
01:11:25	Sam Parr	I mean, yeah, that sounds good to me, too.
01:11:27	Jason Lemkin	I don't know how everyone's like... There's obviously an element of privilege in this, right? For folks who are living in different geographies, others will think this is disgusting, and I think they're right. But I also think folks who have made a little bit of money in tech, or otherwise, are happy to live on relatively low amounts of money with their savings or others, rather than just take a job no matter what. It's what the internet says, right?
01:11:49	Shaan Puri	"Jason, is this a stupid mental model? I do this sometimes. If I want to hire somebody and I'm interested — say they're $200,000 a year — and my business has a 20% net profit margin, I think to myself: this person basically has to generate $1,000,000 of extra revenue just for me to break even on their salary. Now, maybe they're doing it in coordination with the team, but the addition of this person, for it to be a positive ROI, I have to believe that adding them yields me $1,000,000 of extra revenue. When I run that calculation, 90% of the time I'm like, "There's no chance. No — this wouldn't even make sense here." So I don't know if my mental model is broken or if I'm actually onto something. Is that a really dumb way of thinking about it?"
01:12:33	Jason Lemkin	That is the model for sales reps. Traditionally, a sales executive—whether they sold software or an automobile—on base and bonus would take home 20% of the profits. Car dealers have low margins, right? But if you look at the margin to a car dealer, the salesperson selling you the new Lexus, net‑net, would take about 20% of the dealer's profit. If the dealer made no profit, they'd make nothing. If they ripped you off and got you to pay list plus the CD player for $1,000, they'd make a lot. Sales is the same way because in software the margins are almost 100%. Reps would traditionally take home 20–25% of what they closed. For other sales roles, it's really just about leverage, and you're often just hoping to break even.
01:13:17	Shaan Puri	But even to break even—let's say it's a $200,000 employer—to break even on that you need $1,000,000 of incremental revenue.
01:13:23	Jason Lemkin	Yes. But one way—this is what I learned: law firms were traditionally structured in a slightly interesting way, and it may be different today. Traditionally, most law firms—not the big litigators, but your corporate law firms that help you on transactional stuff—never made money on the associates. I didn't know this until I asked a partner I worked with. He said: > "Because I just don't wanna do that work anymore. I don't wanna write the certificate of incorporation for Sam. I don't wanna do the NDA. Like, we don't make any money—net, net, net—of training and offboarding and churn and the desk and the computer. We don't make any. But I just don't wanna do the work." For your organization, you'll also hire just to not have to do the work, right? But the problem with AI is it lets us do more of the work. That's the weird dislocation. Instead of paying that person—you can't even find the $200,000 person at some [firms]—you'll just do the work yourself with AI. That's the cognitive load. But that's why it's the mix: I don't think everyone can be profitable. Some folks just have to reduce the load on everybody else, but AI is gonna do more of it, yeah, right.
01:14:22	Sam Parr	You're full of information. I— you know, I think Theo Von described one of his guests as a "blind Russell Terrier" or "Jack Russell Terrier"—or no, sorry, a "deaf"... What was it? A deaf— I fucking ruined this joke, but.
01:14:37	Jason Lemkin	"We're gonna laugh anyway."
01:14:39	Sam Parr	What was it?</FormattedResponse>
01:14:40	Jason Lemkin	"Deaf."
01:14:42	Sam Parr	Jack Russell Terrier — where he gets out of the car and goes crazy to try to stop him from yapping. Yeah, that's kind of like what you are. It's like we just hit record and you just roll; we're on board and listening to you. You're the man. You're just full of, I think, really good comments that are made in strong taste and tend to be right.
01:15:07	Shaan Puri	Jason, don't get confused. When he calls you a "blind, deaf dog," he's trying to give you a compliment; he just does it a little bit differently, yeah.
01:15:15	Sam Parr	Yeah, and grumpy.
01:15:16	Shaan Puri	Yeah, Andy Glodger. Grumpy. So you're welcome, and welcome to the show. There's a goodie bag for you on the way out.
01:15:23	Sam Parr	Thanks, man. You're the best. That's it — that's the pod.

Discover

Listen

Catalog

Guides

Connect

How the smartest founders are winning in AI

Transcript