← The SaaS Podcast - Growing Profitable AI SaaS & AI Agents
The SaaS Podcast - Growing Profitable AI SaaS & AI Agents
450: Lindy: Pivoting a $50M Startup to Build AI Agents - with Flo Crivello
00:45:55
Loading summary
Transcript89 lines
- [00:00]
Omar Khan
Foreign welcome to another episode of the SaaS podcast. I'm your host, Omar Khan, and this is a show where I interview proven founders and industry experts who share their stories, strategies, and insights to help you build, launch, and grow your SaaS business. In this episode, I talk to Flo Crivello, the founder and CEO of Linde, a no code platform for building AI agents to automate workflows. In 2020, Flo was running Team Flow, a virtual office startup that raised over $50 million and grew to over 50 employees during the pandemic. But when people returned to the office, growth completely flatlined. With no path forward, Flo was forced to pivot. He had to let go of two thirds of his team, which was one of the most painful decisions he's made as a founder. The idea for Lindy came from a simple request. His sales team was spending hours updating Salesforce after meetings and asked if AI could do it automatically. So Flo started building a tool to solve the problem. He kept making it more general. First, it could update any field in Salesforce, then any CRM, then any tool. That's when he realized he was building something bigger. An AI agent platform. So in March 2023, he launched Lindy with a demo video to build his waitlist. But the product was terrible. For example, when users tried to send recruiting emails, Lindy would literally write the user wants me to send an email to 50 engineers and send that to the candidates. But early adopters were surprisingly forgiving. They were buying into the vision more than the actual product. And over the next several months, Flow kept iterating and slowly built it up to over 100k in MRR. But despite the early traction, Flo knew the product still wasn't working like it should and was falling way short of what customers actually needed. So he spent five or six months rebuilding everything from scratch. The breakthrough actually came when they repositioned Lindy as Zapier for AI. Suddenly, people understood what it was first. Five days after quietly launching version 2.0, a YouTuber discovered it and created a video calling it the best AI agent platform he'd tried. That video triggered explosive growth. Today, Lindy generates high seven figures in ARR with just 25 employees. In this episode, you'll learn how tweeting 20 times a day for years helped Flow get 70,000 waitlist signups from a single demo video. How? Flow's sales team asking for Salesforce automation led to discovering a much bigger opportunity in AI agents.
- [02:47]
Flo Crivello
Why?
- [02:47]
Omar Khan
Having 100k in MRR wasn't enough. To stop him from rebuilding the entire product from scratch. We talk about what happens when you launch a product so broken it sends nonsense emails like to job candidates and why users were surprisingly forgiving and how changing their positioning to a zapier for AI instantly made the product click for customers. So I hope you enjoy. When was the last time you verified your app's emails actually reached users? Someone signs up, expects their welcome email nothing. They try password reset, silence. They assume your product's broken when really those emails are sitting in spam. Mailtrap is an email delivery platform built for product companies that send at scale, trusted by teams at Atlassian, Adobe and calendly get faster delivery, high inboxing rates and industry best analytics. And right now you'll get 20% off all plans with the code THESASSPodcast@MAILTRAPIO. That's MAILTRAPIO. Hey Flo, welcome to the show.
- [03:48]
Flo Crivello
Hey, yeah, thanks for having me, Omar.
- [03:50]
Omar Khan
My pleasure. Do you have a favorite quote? Something that inspires or motivates you?
- [03:53]
Flo Crivello
It's going to sound cheesy, but I really like recent chill chills. If you're going through hell, keep going.
- [04:01]
Omar Khan
A great one for any startup founder. So for people who aren't familiar with Lindy, can you tell us what does the product do, who is it for, and what's the main problem you're helping to solve?
- [04:13]
Flo Crivello
We are a no code platform helping you build AI agents to automate workflows. The problem it solves. Think of any AI agent as like an AI intern. Like soon it will be an AI employee that can do anything for you. For now, it's an AI intern that can do like simple things for you. Some of the things that people find it the most useful for are sales. So like prospecting, outreaching, qualifying leads, support. Like you can deploy like a chatbot on your website. You can respond automatically to your support emails, like check all your statuses and so forth. And personal assistance like meeting scheduling, meeting recording, meeting prepping and so forth. We're typically used by SMBs, so startups call it like 5 to 100 people, mostly in tech.
- [05:01]
Omar Khan
Cool. And give us a sense of the size of the business. Where are you in terms of revenue, customers, size of team?
- [05:07]
Flo Crivello
Yeah, well, like high seven figures of, of of revenue. We are 35 employees plus like five or eight contractors.
- [05:16]
Omar Khan
And you have raised. You're just telling me 52 million, right?
- [05:20]
Flo Crivello
That's correct.
- [05:21]
Omar Khan
Okay, great. So we'll talk about like where the idea came from for Lindy, but maybe let's Just talk a little bit about your background before then. So, you know, at one point you were working as, you know, product guy at Uber and then you initially launched a product called TeamFlow.
- [05:44]
Flo Crivello
TeamFlow, that's right, yeah.
- [05:46]
Omar Khan
So tell us about that. What did the product do and what led you to sort of starting that business?
- [05:52]
Flo Crivello
Yeah, so after Uber, I took a break for six months. The pandemic broke out. So I was like stuck at home like everybody else. And I'm an adult. And so what I do when I'm stuck at home is I code and build stuff. And I started building this thing. And at first it was not even because of the pandemic, it was just a cool idea. I was like, oh, imagine like a spatial zoom, so it's like you can see your video in like a bubble, in like a circle, on like a two dimensional canvas. Like think almost like Figma or like the Sims, because you could build like the canvas yourself and you can move around on that canvas like in a video game, and you can hear and be heard by people around you. So then people can dial in to this space. It's like a virtual space where you can hang out with people and all of that stuff. And so I was just hacking that. Initially it wasn't even as a business idea. It was like, oh, this is so cool and they're going to hang out in this space with my friends. And then it dawned on me. I was like, wait, the pandemic is a thing? Wait a minute, I hadn't thought of that. And every team was remote and I knew that remote work was awful. It was just really hard to collaborate with remote work. So that's when I saw the angle. I was like, ah, like remote work. This is like a virtual office. And things were going really well at first because of the pandemic. So everybody was going remote, everybody was looking for a solution. So we took off really rapidly. And then people returned to the office and you could literally see the waves of COVID you know, this graph. Like there were like multiple waves of COVID You could literally see the waves of COVID on our growth chart. It was just like you could overlap them, it was the same curve. And then when Covid went down and people basically returned to the office, the growth slowed down massively. Like we were basically flat. And so that's when we realized that there was no market there and we pivoted to Lindy.
- [07:38]
Omar Khan
Right, so tell me about where the idea came from for Lindy initially.
- [07:44]
Flo Crivello
So I was always super into AI, like before even GPT2. When GPT2 came out, I was so super excited about it. And then GPT3 and then the GPT3 API came out. And because we had this, effectively, it was a video conferencing tool, we built a meeting recorder and we used GPT3 to summarize the meetings that were recorded. And then the sales team came to me and they were like, flo, I would love if this thing could not just summarize my meetings, but also update my Salesforce for me. I hate updating my Salesforce every day. It takes me so much time. And we were like, sure, I can't say why not. And so we used the GPT3 API to just extract the stuff to update the Salesforce entity with. And then little by little, we just built it in a more and more and more generalizable fashion. So at first it's like, okay, it's Salesforce. Now it can be any field on Salesforce. Oh, but now we're going to release it to customers and some of them don't use Salesforce, they use HubSpot. So that can be any field on Salesforce or HubSpot, and that can be any field on any CRM. Oh, what if it could also create tasks on linear. And so we just keep climbing up this ladder of abstraction until we're like, wait a minute. And we built this framework and it was before ChatGPT. We built this framework and we realized, like, wait a minute, LLMs can actually do stuff. Not just extract text or summarize text or generate stuff. They can do stuff. And so that's when this vision of the AI employee came together. And then ChatGPT came out, LangChain came out. We were actually in very, very early touch with Harrison Chase from LangChain, when it was just him and like a GitHub library, it still had a job. And so like early 2023, we decided to halt pivot to Lindy and man, look, we were like a team of 50 plus at the time. I had to let go of two sales of the team. I had to fire the entire go to market team, which was ridiculously painful and one of the most painful parts of my career as a founder. And we went down and started to build this new vision.
- [09:44]
Omar Khan
So what was the landscape like at that point? I mean, today you go online if you can, like when the Internet is locked down like we were talking about today, and it kind of seems like overnight everybody is talking about they build AI agents, right? It's this term that you just can't going to get away from. What was the landscape like when you initially pivoted and started, Lindy?
- [10:16]
Flo Crivello
So people weren't really talking of AI agents back then. People were talking, if you remember, of generative AI. It was like it was generating text, it was generating images. People were not talking yet about video. It was not generating videos yet. And so I was like, look, this is all good, but the GDP is not made of copywriters or illustrators. It's made of work and actions. So I wanted to build agentic AI, not generative AI. And the landscape was Adept, which since then has been acquired by Amazon and LangChain. That's basically it. And then there was like Llama index a little bit, but that was basically it.
- [10:51]
Omar Khan
Okay, great. So tell me a little bit about it. So you've kind of built this product. You talk mostly about doing sort of meeting, summarization, updating some of tools like CRM. How did you sort of go about launching the product and finding those initial customers?
- [11:16]
Flo Crivello
We put out a demo, which at first was almost more like a vision statement. So it was like, hey, well, Lindy, we're working on this new cool thing. Here's what it's looking like. And we had prototypes. Some of that was like Figma prototypes, some of that was working prototypes. And we released the video on Twitter, and I think that was March 2023, and we acquired a wait list of 70,000 people. That's how we started in the very beginning.
- [11:45]
Omar Khan
So let's talk a little bit about that, because you've built quite a following on Twitter or X. And a lot of that was before you started, Lindy, right?
- [11:54]
Flo Crivello
Yes, I've got a little bit of a following now. Nothing crazy, but I always tell people two things. One, if you're not on Twitter, you are missing out on the biggest party in human history. You have to be on Twitter. It is my biggest career asset apart from my company. But before my company, it was my biggest. If you'd asked me, you have a choice between quitting your job and wiping your LinkedIn. You never were at your job, which was. Uber was also like a good company. Oh, destroying your Twitter account. I would have picked to destroy my job experience. Like, no doubt, no question. Easy, easy. And then the second question, and the second thing I always tell people is they ask me, how do you go about building the Twitter following? And I always say, it's a little bit like enlightenment. It's like if you seek it, you cannot get it. It's. You don't overthink it. Just tweet all day. Go after volume, go after quantity. More than quality. There is no such thing as a tweet with negative views. So by definition, every tweet gets you more attention. And just tweet all day. Tweet about anything that crosses your mind. Anytime you are in a meeting, anytime you hear something interesting or something funny, just tweet about it and embrace the fact that you are going to get it wrong. And it's going to take you six or 12 months to find your voice. You're going to be cringe and bad for the first six or 12 months, and it does not matter. Nobody gives a shit. Nobody cares about you. The worst thing that can happen is that you get no visibility. It's not even that people will forget about the bad stuff that they saw about you. It's that they literally won't even see it. They literally won't even be aware of it. So the only world where people are aware of you is a world where you're good.
- [13:32]
Omar Khan
Yeah, that's a good way to think about it. So when you say tweet all day, what are we talking about? How frequently were you or how many times a day were you posting?
- [13:46]
Flo Crivello
I actually once had a script that was running on my server that would every day send me a notification, like an email saying, hey, you've tweeted X times today. And My objective was 20 times a day. So that gives you a feel. And it sounds like a lot, but some of these tweets are like brain farts. My most successful tweets, that's another thing I've learned is there is this quote I love that's like, nobody knows anything. It's like a very famous movie star or director who spent his career making movies, and it's like, no one knows anything. You've spent 30 years of your life making movies. You still don't know whether a movie is going to work out or not. So it's just optimized for number of swings of the bat is what Marco Rosen calls it. And so some of my most successful tweets, like, there was one of them in particular. I was literally in vacation with my girlfriend. I was on a beach in Hawaii. We were like, wrapping up. There was like a towel on the beach. She was like, all right, let's go home. Like, okay, sounds good. Oh, I just had an idea for a funny tweet, and literally she was like, wrapping up the towel. I pick up my phone, I tweet the tweet. It took me 30 seconds. It was a brain fault. Like, ah, this is funny. All right, let me finish wrapping up. That tweet went crazy viral. Was one of my most successful tweets ever. It's just brain farts. Just tweet anything that's on your mind.
- [14:56]
Omar Khan
And do you think for somebody who. Who is on Twitter or X today and doesn't have a lot of followers, hasn't tweeted a lot, and maybe they've been on the platform for a few years, but they haven't done much with it, Is it still worth them investing time to do that?
- [15:18]
Flo Crivello
Absolutely. There's new accounts coming out all the time that in the span of six months pick up to hundreds of thousands of followers. I will say that the cold start problem is real. I honestly, I've been on Twitter like 17 plus years at this point, so I don't really remember how I cold started my account. I think the typical piece of advice that people give is be like the reply guy. And you want to be mindful about it and stuff. But yeah, if you bring value to people in people's replies, and I think the best way to bring value is to be funny. Honestly, just be funny. Then you can call, start your account.
- [15:56]
Omar Khan
Yeah, cool. Okay, great. So you've obviously invested years building this audience. What was the. It was just like a video you posted of the kind of like, what Lindy was about or the vision for this business?
- [16:14]
Flo Crivello
It was just a video. It was like me talking to a camera. Again. I think one thing that you have to be on Twitter is authentic, Just very authentically. Me talking to a camera. I was like, hey, I'm Flo, I'm following Lindy. This is the product. It's pretty sweet. Sign up to the waitlist.
- [16:30]
Omar Khan
Okay, great. So that sounds like a super promising start. And then were you charging for the product, like from day one?
- [16:40]
Flo Crivello
Yeah.
- [16:41]
Omar Khan
And did you. I forgot to ask you, did you raise any money? Like, at what point did you raise the first. Your first round?
- [16:48]
Flo Crivello
Well, well, it's complicated because Lindi is a pivot from TeamFlow, so it's the same legal entity. So we'd raised a lot of that money for TeamFlow.
- [16:54]
Omar Khan
How much did you raise for TeamFlow?
- [16:56]
Flo Crivello
Well, 50 plus million. Oh, wow.
- [16:58]
Omar Khan
Okay. So you raised most of that at that point? Yes, got it. Okay, cool. So, all right, so, well, great. But there's also a lot of pressure comes with raising that kind of money.
- [17:09]
Flo Crivello
Right. But look, the pressure comes from within. It's not like the investors are calling me every day. It's like, where's my money? It's like, look, if you've raised money, you've got a obligation to your shareholders. And I think frankly, your first obligation is to not spend it until you've got a good use for the capital. Until you know that you can deploy it effectively, don't spend it. And so that's what I did for tflow. We raised a lot of money and I didn't spend it. And to this day, we've got the vast majority of the money in the bank. And the investors were grateful. They're like, yeah, thanks for not spending the money until you could put it to good use.
- [17:36]
Omar Khan
So what happened next? So you got these people on the wait list. You launch the product. What happens?
- [17:43]
Flo Crivello
Well, we build, we start taking people off the waitlist, we invite them to the product. It goes wrong, the product doesn't work. It's not easy enough to use, it doesn't retain. And so from there it's just iteration, just iteration, iteration, iteration. Like, look, the YC adage is really good and true. It's just build something people want. And the way you do that is, is you talk to your users and you do what they tell you to do, basically. And it's that simple. And I think the trap not to fall into here is like, by the way, it's not even a trap. Sometimes it works. It's like, I do still think you should have a vision, so you should have like a weak inductive bias towards a certain direction. I call it like climbing the Everest. So like, you want to go to the top of the top, you're looking for base camp along the way. You can't go straight to the top. If you find a base camp that's like on some other mountain, you don't go there. It's not where you want to go. Some people do and it works out for them. It's like the whole lean startup thing, it's like, all right, fuck the vision, just do whatever the market wants. I think that's a miserable existence. I think as a founder, you need to find meaning in what you do. And so I think you do need to that vision. And so, yeah, just talk to your users and build what they need.
- [19:05]
Omar Khan
I talk to a lot of founders who are, when they're at that stage, they're reluctant to ship the product. And the argument that I hear is, well, I'm going to get one shot at building credibility with my market. And so I need to make sure that this is right. And, and what happens inevitably is this Thing drags on for months and months. So just to reassure maybe somebody who's in that situation, who maybe thinks the world is going to end if their first version of the product is shit, why is that not something a big deal as it may seem?
- [19:46]
Flo Crivello
I really love what's his name, Reid Hoffman, the LinkedIn founder. That quote from him that goes, if you're not embarrassed by the first version of your product, you launch too late. You want to confront your idea to reality asap. Like, I don't know anything about your idea. All I know about it is that it's wrong. And that's all you can know about it. It's wrong. And so the only way you want to find out how wrong is in what way is app. And the way you do that is by shipping it and confronting it to reality. That's it. If you look at the V1, which you can find, you look at the V1 of Uber online or Airbnb or Facebook or Reddit or any of these websites, it was embarrassingly bad. It was barely working and nobody remembers. Nobody gives a. I think people greatly overestimate how much of a shit other people give about that. I forgot who it is who said, I think it's Naval Ravikant who said the difference between the first time founder and the second time founder is the first time founder has got that mindset. He basically operates under the assumption that people care about them and so they're like hush hush about their idea. That's one thing that my pet peeve is like founders who are in stealth mode also, by the way, I think it's ridiculous to be in stealth mode in public. But it's ridiculous, let alone at a dinner table. It's like, bro, there's like six of us. We're not going to steal your idea. We're very busy. Just tell us. And probably your idea is wrong anyway. So just tell us about your idea so we can give you feedback. It's like, I can't tell you my idea is like this precious little stuff. Like you have no idea how genius it is. It's like if your idea is what protects you, then you're in bigger trouble anyway. So anyway, yeah, just look at these startups. Assume that nobody cares about you and B is the second time founder who basically appreciates that nobody cares about them. Startups die by not making noise. No, startups die by making too much noise. Just make as much noise as possible. Cry your idea off the rooftops and hope that someone gives somewhere.
- [21:41]
Omar Khan
When you said the product Was not, not very good. Can you give me an example of what kind of problems users were experiencing?
- [21:50]
Flo Crivello
Oh, I mean, it was embarrassing. It was like the agent would. We gave this demo that was like, look, I can recruit with my AI agent. So I'm going to ask my AI agent, like, find me 50 software engineers in San Francisco. And then it finds the software engineers and then you're like, cool, now send them an email to the them to join my startup. So then it sends them an email and people would try that. And it would send an email to 50 software engineers too. It's a lot from you. And it would say something like the email body would say something. The user wants me to send an email to 50 software engineers. That's the email that people received.
- [22:28]
Omar Khan
So what was happening? Were those initial customers, were they people you knew? They were just random people who, who saw the post on Twitter and how forgiving were they when things like that happen with the product?
- [22:42]
Flo Crivello
Surprisingly forgiving. Very forgiving. Look, I think people self select, they know they're signing up to a random demo that they saw on Twitter. So the vast majority of these customers were not people I knew. Almost disappointingly so, actually. I could barely get my own friends to sign up to this thing. Why won't you use my product? And yeah, I mean, like when these things would happen, look, yeah, sometimes they're pissed, you know, but my biggest fear was actually perhaps selfishly, that they would post about it on Twitter. And as far as I know, no one ever did ever. Like, we've got many users, like many figures of many users. To my knowledge, that's never happened. There's been one guy who posted on Twitter that he was mad that we've got money. You're going to get those guys. But, but no, they're surprisingly forgiving. It's like I was mortified. I was like, oh my God, I'm so sorry, let me refund you. And they were like, no, don't worry, it's fine. Hahaha. It's funny. Yeah.
- [23:39]
Omar Khan
When we were talking earlier, you also said there was a vision gap, that when people would come to Lindy AI, they would have a certain idea in their mind of what they were going to get. And when they actually signed up for.
- [24:00]
Flo Crivello
The product.
- [24:02]
Omar Khan
It was very different. Right. So aside from the problems, what was the gap between the vision and the reality of the product at that time?
- [24:11]
Flo Crivello
Well, the vision's always been to build an AI employee. And that's compelling in and of itself. Like, even for investors. It's like Duh, yeah, if you can do it, which that's the if people will want it. There's no if of will people want it. It's just pure execution risk. And it's funny the extent to which people buy the vision, not just investors, people sometimes also buy the vision, meaning they literally buy it. They give you money for the vision. They're like, ah, this is the vision, this is the product. And then they're like, here, take a couple hundred bucks a month. And I want to. I want to see what happens. I want to experience the early stages of this vision, which I think is also my point about why are people so forgiving? Because they know what they're signing up for again and they self select there. It's like very early cohort of very early adopters. They're really pioneers. And that's the cohort that you want and that's the cohort you're going to get naturally. You don't need to worry about it. That's the cohort that if you're like a mechanic, you can give it like a new piece of a new engine that you've designed and it's oid. It's disgusting, it's half broken, but that's the code that can get their hands dirty in touch your piece of that engine and be like, oh, yeah, I can see it. I can see it through the slime and this guessing grossness and brokenness of it all. And I can see what you're going for here. Yeah, so there was this gap and I think the fact that people bought that vision gave us the rope that we needed to carry through and took us a year and a half or two years.
- [25:37]
Omar Khan
How did you position the product two years ago? Doesn't sound like a long time, but in the world of AI, it probably is. And when you sort of describe it in terms of workflow, what do people know? They might know zapier. And so some people might assume it's kind of like that. Other people may kind of hear the vision and think it's something much more than it actually is today. So how easy or hard was the kind of the positioning and kind of having clear messaging for people in terms of what Lindy was and where it fit in the market.
- [26:25]
Flo Crivello
I wouldn't call it hard because once you understand the framework, it's not that hard. So basically the framework, there's a book about this that's called the battle for your mind. I believe the author's name is Al Rising, and the analogy he uses for positioning is you want to position against what already exists in the prospect's mind. So he uses this analogy of Velcro. It's like Velcro, the way it looks and works if you zoom in is you've got hooks and loops. You've got hooks that just grab onto the loops and that's how it sticks. And so it's like there's these two sides and the hooks exist in the prospect's mind. There are things that are already there. You cannot position in a vacuum. You've got to be mindful of what already exists there. And. And you've got to design your loops around these hooks. You can't just be this new thing. And what we did, and from a positioning standpoint, what started working is we started to position actually ourselves against Zapier. There's this other framework from April Dunfield that's also called positioning that I really recommend, and she does recommend basically picking your positioning in a way that puts you in the best light or following. First positioning was AI employee. Why does it put us in the best light? Well, employee already exists. You know what an employee is, and it puts you in the best light because it's so obvious in which way an AI is better than a human employee. It's like a hundred x faster, better and all of that stuff. The reason why that was a bad positioning is because the product was just falling too short of that. So the second positioning that ended up really working was the pure of AI. We had the tagline that was like if Zapier and chatgpt had a baby. And then we ended up retiring that tagline because we learned. To my surprise, I thought Zapier was like a household brand name. But no, actually a lot of people would join our website and be like, what the is Zapier? Oh, we're actually advertising our biggest competitor. So, okay, never mind. But I call it the Notion head fake. Like Ivan, I think calls it like a chocolate covered broccoli. Where notion. Obviously immensely successful company, but it took them a while to be successful. And Notion 1.0 was not successful because it was really putting front and center that vision they had of these LEGO bricks, these building blocks that you can use to build whatever software you want. So it was this weird alien new thing that's hard to understand. Notion 2.0 was a note taker. It was a wiki. You sign up, you see a blinking cursor, you can take notes like, oh my God, you know what it is? It's a note taker. Notion is not a note Taker. Ivan never set out. That was not his vision to build a notetaker. Right. He wanted to build a no code app platform, but by pretending it was a note taker, it made it so much more accessible. And he told me he was like, many founders resist the comparison to things that exist because their startup is this one beautiful, unique snowflake. But no, he says you should embrace the comparison. Like familiar is good, it's alien. That's bad, it's confusing, that's bad. And being familiar does let you tap into these hooks that exist in people's minds. So Lindy 2.0 Z, pure of AI and even the product looked very similar to ZPO. All of a sudden the 1.0 was more of an alien product, like a new type of thing, like this agent builder. The 2.0 is the current incarnation of the product is this thing where you have these boxes and arrows under the hood. It's a very different thing, actually. It's not a workflow dimension platform, it's an agent builder. But it looks very similar.
- [29:58]
Omar Khan
You've got the 30 users, it's not working quite like it's supposed to. You're having these conversations, you're improving the product, you're iterating, all that stuff. And I want to kind of talk about what happened last year when sort of growth really took off, but just kind of describe to me what was going on up until that point. Up until about May last year.
- [30:31]
Flo Crivello
How.
- [30:31]
Omar Khan
Much growth were you seeing? What did churn look like until the 2.0?
- [30:37]
Flo Crivello
Until things started working, it was like middling growth. It was really just like chugging along painfully and you could tell things weren't working. I really love this quote. I think it's from Marc Morrison, I forgot who said it. It's like until you have product market fit, it's like pushing a boulder uphill and after you have it, it's like chasing the boulder downhill. And I felt I misunderstood the quote because until you have product market fit, you think product market fit is the holy grail. Well, actually it's a pretty mundane thing. It's like every successful business out there has product market fit and there's a lot of successful businesses out there. So it's like, ah, the holy grail. And once you have it, you're chasing the border downhill and then it's forever after everything is perfect. It's like. No, actually chasing the border downhill is also hard. It's a different type of hard. It's a much better type of hard, but it's Hard. It's really hard. And we felt it almost overnight once we had it. And it felt like chasing a bulldoer downhill. It was like, oh my God, it's overwhelming. Everything is breaking. We have more customers than we can handle. The soles are breaking. Like Google with exactly. With exactly. We're not going to bed tonight. It's like, it's crazy. It's overwhelming. And so for us, that happened once we released Teamflow, sorry, Lindy 2.0.
- [31:45]
Omar Khan
And that was triggered by somebody who kind of came to you back about five, six months earlier, asking you how they wanted to use or at least telling you how they wanted to use Lindy.
- [32:01]
Flo Crivello
Yeah. So we had this moment at some point where the idea of Lindy 2.0 and of ZPR, of AI started blooming in our minds. So we started deciding to rebuild the product. We rebuilt it, we released it, and at first we decided not to announce it, which was a mistake. Again, you should make as much noise as possible. You should announce everything possible. We decided not to announce it. We decided to first quietly put it in the hands of users. Which even when you say it like this is like such weak cheats. Like, oh, it's just quietly. It's like, no, let's just make a splash all the time. So quietly put it in the hands of users. And five days after we quietly put it in the hands of users, Matt VitPro, he's this YouTube influencer whom I am so very grateful for, picked it up. He was one of these customers, was paying more for the vision than for the actual product because he could sort of see it. And he was like, he tried Lindi 2.0 and he was like, oh my God, this is awesome. He really liked the new version. And he made a video about it where he was. So he was extolling the new product, he was like, oh my God, this is like, I've tried all of them. This is by far the best these guys have been cooking. And then things really started taking off. And at first we thought it'd just be like a one time spike, but actually it wasn't. It's just like inflected and forever continued on this trajectory. And it's actually kept accelerating since then.
- [33:20]
Omar Khan
But what drove you to rebuild the product?
- [33:23]
Flo Crivello
It wasn't working. So there was one time in particular where we had a big prospective customer that we were talking to. We were talking to customers all day, all the time. And there was this one that gave us a list of the exact use cases that they wanted. And it clicked at that moment because that's when we realized like, okay, this is actually pretty similar to what we've heard from this and this and that and that and that other customer and the product today cannot do it. And most importantly the product, the current plan will not allow the current product to do it. We're very far from that. I think for a very long time we were overly cavalier. We were like, ah, the model will get better. And we were like, no, no, no, no, no. This guy needs this to work every single time. And the current paradigm is broken. We need to rethink the paradigm that the product is built over. So that's when it clicked.
- [34:17]
Omar Khan
And the 2.0 took you about five or six months to build. Yeah, that's a big investment to make. I mean you said nothing was working, but you still had revenue. You were still doing what, over like 100k mrr around that time?
- [34:32]
Flo Crivello
Yeah, little about.
- [34:34]
Omar Khan
So it's not like, you know, you have no customers and you can do whatever you like. There's still some level of risk and a pretty long time to spend building something not knowing whether it's going to.
- [34:47]
Flo Crivello
I think that's something, I guess I'm doing a bit different. And I've received positive feedback about it from investors. Perhaps surprisingly. At first I thought they were joking in a bad way. They were like, ah, flu. Like, I guess you're never afraid of just destroying your entire existing business, are you? And I was like, I'm sorry. They're like, no, no, it's good because look, nobody is in this game to make a million bucks a year. I'm not in this game. Investors are not in this game. Maybe, maybe some founders who are listening to us are in this game. You shouldn't be in this game. In a world where like a billion dollars a year businesses exist, why would you build a one million bucks a year business? You know, you should go after the billion bucks a year business. Like absolutely. And so I always knew that like the 99.9% of our revenue existed in our future, not our present. And so eyes on the prize. So yeah, we had this existing customer base. First of all, you can look at the metrics, but most importantly, you feel it in your guts. You just know it's not working. And I was like, yeah, this is not working. And this new vision and version is so much more exciting to all of us and we feel a lot more optimistic about it. So yeah, we're going to do that.
- [36:00]
Omar Khan
And so that was around, I think you said October time And then growth just took off in large part in thanks to your YouTuber friend, Matt, who you don't. Do you even know him? He was just like some random dude, right, who created that video.
- [36:20]
Flo Crivello
I don't know him. I don't even know if I've met him. I think I should. I think we sent that email like, hey, thanks, Matt, but we should have done a podcast with him or something. We should have said like a bucks of chocolates at least or something. Yeah.
- [36:32]
Omar Khan
And then, so, okay, so things take off and now you are feeling like you're chasing that boulder downhill. It's growing. And now in many ways, I've had a founder describe it to me. Like, the train is now really moving fast, but it still feels like the wheels could come off at any minute.
- [36:53]
Flo Crivello
Oh, yeah, big time. And they do come off. They do come off. Yeah. Just yesterday morning, we had one of these moments where we're like, the wheels are coming off. And actually it was a false alert. But 8am Some member of the team sends a message like, hey, guys, there's this really bad business situation. This metric is way down. A business metric is way down week over week. We need to jump on it right now. And so one room, everybody jumps in a room together. And so it's like, okay, step one, there's basically understand and do. There's two buckets. So understand Jack is his name. Go look at the data. I want to understand the funnel. I want to understand how the conversion has been training over time, like A, B, C, D, E. In the meantime, we don't have time to wait for Jack to come back, but do come back to us in half an hour or one hour. And you go and do. We're going to change this and this and that in the funnel right there. And it turns out it was a false alert. Jack got back to us still, guys, everything's fine. But anyway, it does feel like the wheels are going to come out all the time.
- [37:54]
Omar Khan
The other thing that happened was suddenly you had a lot of competitors kind of come out into that space. Right. I mean, there have been products out there who are doing some similar things to Lindy. But then also, even people who aren't building a product similar to Lindy are using messaging about AI agents that they. And so suddenly things get very confusing, I think in the market, at least for the end user who's trying to decide, what the hell do I use?
- [38:31]
Flo Crivello
Yeah, I think it's surprising because I compare it to the universe. You know, how there's literally tens of trillions of trillions of stars in the universe, and yet the universe is 99.99% empty space. I think it's the same thing here. It's like there's a lot of AI agents, platform out there, and yet when we talk to customers and when we talk to the market and to our users, we don't see them. So again, I think here the cliche answer is also true. It's like, focus on your business, don't focus on your customers.
- [39:01]
Omar Khan
Yeah, I think that's great. All right, we should wrap up. Let's get onto the lightning round. I've got seven quick fire questions for you. Ready? What's one of the best pieces of business advice you've received?
- [39:11]
Flo Crivello
Build something people want.
- [39:13]
Omar Khan
What book would you recommend to our audience, and why?
- [39:16]
Flo Crivello
Atlas Shrugged by Ayn Rand. Every human should read Atlas Rugged. It's been the most foundational book for me and my personal philosophy. I wouldn't be here in life or even geographically. I wouldn't be in the US if it wasn't for this book. Why do I recommend it? I recommend it because it puts forth this vision of man as. And man's mind as a creative force. And it depicts in visceral detail the forces that stand in the way of that force, that creative force.
- [39:49]
Omar Khan
What's one attribute or characteristic in your mind of a successful founder?
- [39:53]
Flo Crivello
Unstoppability.
- [39:54]
Omar Khan
What's your favorite personal productivity tool or habit things?
- [39:59]
Flo Crivello
The app is like, I've been using it for 20 years now, and the habit for me is don't confuse goals and tasks. People are always like, there's like this to do list. I don't care about the to do list. Like the to go list. I almost want to call it right. So I make it a point to have my goals in my to do list. And I try to look at my goals more than I look at my tasks.
- [40:22]
Omar Khan
That's nice. I love things as well. I use it. What's a new or crazy business idea you'd love to pursue if you had the time?
- [40:28]
Flo Crivello
I would love. I really love the whole toddler CD movement. Balaji Srinivasan, who's a friend of mine, had this crazy idea that I keep thinking about. It's like, let's buy some cruise ships or lease some cruise ships and put them in international waters and let's build a city on the water. I, to this day, don't understand why that would not be possible.
- [40:49]
Omar Khan
What's an interesting or fun fact about you that most people don't know?
- [40:53]
Flo Crivello
I Am from a quite poor background and I moved here to the US 12 years ago after reading Atlas Rugged with basically no money, no plan, no job, no network. And I spoke actually very poor English back then, so I had to also ramp that up.
- [41:10]
Omar Khan
And finally, what's one of your most important passions outside of your work?
- [41:13]
Flo Crivello
Outside of what? What do you mean, outside of my work? No, I mean, seriously though, I basically work. That's all I do. I think, like, you have time enough in your life for two things just to. And you've got to pick between work, like family, fun, health, I'm sorry to say, learning, friends, that's it. And so me, it's my work and my family. My family is basically my girlfriend and my cats. And I read and I try to exercise.
- [41:47]
Omar Khan
Awesome. Well, thank you for joining me. It's been a pleasure. Thanks for unwrapping the story and sharing some of your experiences along the way. If people want to check out Lindy, they can go to Lindy AI. And if folks want to get in.
- [42:01]
Flo Crivello
Touch with you, Flowindy AI hit me up and axe, you know, a L T I M o r on X.
- [42:09]
Omar Khan
Awesome. Thanks, man. It's been a pleasure and I wish you and the team the best of success.
- [42:14]
Flo Crivello
Thank you so much, Omar.
- [42:15]
Omar Khan
My pleasure. Cheers. Your perfect onboarding flow means nothing if the welcome email never arrives. Too many transactional emails miss the inbox. Lost trials, failed activations, users who can't reset passwords. It adds up fast. Mailtrap is an email delivery platform built for product companies that send at scale, get fast delivery, high inboxing rates and 24. 7 expert support trusted by over 150,000 monthly active users. Try it today at Mailtrap IO and get 20% off all plans with the code THESASSPodcast. That's Mailtrap IO.