🎧🍌 The Future of Software is Humans and AI Working Together | Steven Fabre, Liveblocks
What most people get wrong building AI-native software, how large companies are actually adopting AI today, lessons going from designer to CEO, and how to recover from a co-founder breakup
Everyone is trying to add AI to their software products right now. And according to Steven Fabre, 99% of people are doing it wrong.
Steven’s the co-founder and CEO of Liveblocks, building ready-made AI copilots and collaboration infrastructure for developers.
Steven is one of my smartest friends on how people are actually using AI on a day-to-day basis. We talk about what most people get wrong when building AI software today, why the future of software is humans and AI working together, how to treat it as more than just a copilot that sits on top of your product, and what he’s learned about how large companies are actually using AI right now.
We also talk through Liveblocks journey of evolving from real-time human collaboration components into one that also incorporates AI, what he’s learned going from a designer to a CEO, how to do a successful Product Hunt launch, and how he rebuilt the company after his co-founder stepped away.
Brought to you by Numeral
Numeral is the end-to-end platform for US sales tax and global VAT.
It automates all the manual, headache-inducing parts of staying compliant across thousands of jurisdictions around the world. It integrates with practically every other tool you could use. And they stand behind their work, guaranteeing they’ll reimburse for any penalties or interest charges you come across while using the product.
Try it here and upgrade your sales tax and VAT compliance to the 21st century.
To inquire about sponsoring future episodes, click here.
Timestamps to jump in:
2:17 Liveblocks: Infrastructure for people + AI
6:08 Wrong ways to add AI to software
8:05 Why humans and AI must collaborate
12:35 How AI will change software UI
18:58 AI search optimization
26:20 How to get #1 on Product Hunt
32:33 Liveblocks 1.0 to 3.0 evolution
36:40 Why collaboration software is so hard
38:38 How customers use Liveblocks
42:36 Hiring a coach to get better at sales
47:07 Steven’s biggest enterprise sales mistakes
50:28 How AI changes GTM and funding milestones
57:57 Going from a designer to a CEO
1:01:06 How Liveblocks first started
1:04:56 Importance of design in company building
1:06:51 Learning to become a CEO
1:12:29 When his co-founder left 5 years in
1:15:49 Becoming stronger hiring a new Head of Engineering
1:22:10 Remote culture: what doesn’t work
1:24:08 Remote culture: what does work
1:26:47 Importance of autonomy on remote teams
1:28:05 Most underrated basketball players
1:33:38 ACL injury that kickstarted his first business
Referenced:
Check out Liveblocks
Careers at Liveblocks
Find Steven on X / Twitter and LinkedIn
Related Episodes
👉 Stream on YouTube, Spotify, and Apple
Transcript
Find transcripts of all prior episodes here.
Turner Novak:
Steven, welcome to the show.
Steven Fabre:
Thank you, Turner. Good to be here.
Turner Novak:
Yeah, it’s good to have you. It’s good to... I think I’m going to pronounce your name right in the recording that we just played before this, the little intro segment. One thing I can pronounce, though, is Liveblocks, which is the name of the company that’s pretty self-explanatory. Can you just explain really quick what it is for people who don’t know?
Steven Fabre:
Yeah, sure. Liveblocks is an infrastructure, collaboration infrastructure for people and AI. Think about Figma or Google Docs or Notion. We basically enable you to build collaborative tools, like those tools that I just mentioned. If you want to have like comments, notifications, or real-time editing and see the live cursors, that’s basically what we enable other companies to do. We give you all the building blocks to rebuild those kind of experiences. One big area we’re focused on now is training AI as a collaborator. How can you integrate AI into an existing product so that it does behave like a human would? As you can actually mention the AI and then AI goes on and is able to actually do work like a person would. Yeah, that’s basically what we do at Liveblocks, and yeah, happy to dive in more details if that’s helpful.
Turner Novak:
No, I think that’s good. I think we’ll cut the recording right there. Good episode.
Steven Fabre:
Yes.
Turner Novak:
How does that work in practicality? If I want to incorporate AI into my product, just generally speaking, what does that look like?
Steven Fabre:
Well, generally speaking, it’s very similar to companies like Clerk, for instance, and Clerk, they do authentication. I know you did a podcast with Colin recently. We basically offer a set of pre-built React components for specific collaboration features. If you need a commenting system in your app, we actually have pre-built components for that. You can drop them into your app and quickly add Figma-like or Notion-like or Google Docs-like comments into your product. That’s how you integrate it with it.
In terms of AI, there are different ways you can do this. We do have an AI chat component that you can integrate, but we also have APIs that you can use to integrate AI directly into some of the other parts of your product, commenting or notifications or even enabling the AI to actually change or edit a document. That’s something that you can also do with Liveblocks. That’s a slightly different type of integration, but that’s basically how you would do it. Developers are basically our users. Developers end up using the Liveblocks APIs in their product.
Turner Novak:
Don’t you need to integrate with all these different models and stuff, too? Do you kind of do that? Or how does that play out?
Steven Fabre:
Yeah, so right now you basically have a way to configure what we call AI Copilot inside of Liveblocks. When you do that, you can pass in an API key from either OpenAI, Cloud, or whatever model you want. Then, we handle kind of all the orchestration for that. That’s one way you can do it. One thing we’re focused on now is enabling people to also integrate with more of those like agent products, such as CrewAI, LangChain, NIA I think it’s pronounced. I forgot exactly the name, but-
Turner Novak:
n8n?
Steven Fabre:
... yeah, n8n. Yeah, so that’s also another way you can do it with Liveblocks. There’s different ways you can integrate with that. We actually don’t provide the AI models themselves. We basically provide the interface. Our expertise is around the interface and the end user experience that we provide to enable people to collaborate with AI. You can basically hook that to whatever model you want, whatever agent that you’ve built yourself.
Turner Novak:
It’s kind of almost in a way how Stripe works where it’s like we built the checkout for you, and then there’s all the different payment methods underneath the checkout.
Steven Fabre:
Yeah. It’s very similar actually in that sense. Yeah. Yeah, pretty much. It’s a different vertical, of course, but yeah, it’s an interesting way to think about it.
Turner Novak:
What have you been seeing in terms of how either hopefully your customers, but also, hopefully all these people are your customers, but the way that people are starting to use AI in products beyond just it’s chatbot, ChatGPT, whatever? What are some interesting ways you think that AI is starting to show up in software?
Steven Fabre:
Yeah, that’s a good question because to me, basically I’ve seen a lot of companies... I mean, I’ve seen them. Every company has a ton of money to spend on AI right now, and they’ve invested a lot over the last couple of years. What I tend to see a lot of those companies do is sort of bolting on an AI chat somewhere in a corner or like a very specific part of the app and calling it a day.
Turner Novak:
That’s not good?
Steven Fabre:
I don’t think that’s good, no. I don’t think that’s good. I think this is requiring a lot of people to learn how to interact with this new interface. You have to go to a new part of the app, chat with this AI, and it’s kind of like separate from everything else. I feel like this kind of experience is quite broken, and in many ways, people already used to ChatGPT, Cloud, or whatever model that they’re already using on the site. What it’s like to see, and even myself, I don’t know about you, Turner, but I personally have like a ChatGPT tab, and that’s where I always go to. I’ll do a lot of back-and-forth. I copy data from a product that I’m currently using, paste it in, and sort of resolve that stuff myself. That’s how I collaborate with AI today.
A lot of companies have tried to replicate that model and sort of like bringing a worse version of ChatGPT inside of their product in a little corner. I really don’t think that’s the right UX, and from the people that we’ve talked to, the companies that we’ve talked, they’re not seeing a great ARI on this, to be honest. What I’m seeing now, and it’s something we’re enabling at Liveblocks, and this is something we already use at Liveblocks, is really treating AI like a collaborator. What that means is being able to access AI wherever you’re already doing the work. What that means in practice for us, it’s basically being able to mention AI inside of Slack, for instance.
You may have some feedback from a customer in a shared customer channel, maybe you have feedback internally with somebody on the team. Just being able to add mention an AI engineer, for instance, to be able to fix a bug or fix something that somebody shared with you right from Slack with the right context. Then, having this AI create like a Linear ticket, a Linear issue about this, and then having the AI actually doing the engineering work and submitting a PR for that bug or that issue that you’re solving. To me, that feels more like the way you should approach AI to be successful with AI because you’re not forcing new users to learn a new pattern.
They are using the same patterns they’ve been using for years collaborating with other humans, basically. It just feels natural, and that’s basically what we want to enable companies to do. We want to give them all the tools and APIs to enable those kind of workflows inside of their product. That means being able to mention AI inside of the comments, to give it the feedback about whatever thing you’re commenting on, being able to mention AI in like a text document directly so that the AI is able to actually do work like a human would.
I do feel like connecting AI in that sense in various parts of the app from a collaboration standpoint is the way to go. That’s what I’m starting to see a lot of like world-class companies do. This is something you’re starting to... You’re able to do this with Linear, for instance. That’s really good UX, and I’m seeing a lot more companies kind of following that pattern. That’s what I’m seeing right now. That’s how we use AI efficiently internally, and that’s how we want to enable companies to use AI or enable their users to use AI in their product.
Turner Novak:
It’s almost like instead of you have to... You’re meeting people where they’re at. You don’t have to change an existing use case where you’re almost like instead of mentioning a human, the human engineer on the team like, “Hey, here’s a bug, read this thing. Here’s the GitHub, whatever, click it, read it, go fix it.” It’s actually like, “Hey, just AI, exactly what I just said to the human, but you do it instead.”
Steven Fabre:
Exactly. I really think this is the way to go. We use Devin internally. Devin is this kind of AI engineer. You ping Devin and then Devin is able to actually do the code and then you can review it. That’s a really great way to do it. Another pattern that you tend to see that’s following this model is Linear. Linear has built their own version of agent that you actually create. You can actually ping Linear and then an actual AI engineer can be tagged within Linear. There’s different things that you can enable depending on the workflows and how you want to do your own processes at your company. That’s what I’m saying. You also see like there’s a similar pattern with code reviews as well.
Typically when you build new products, at some point you need to get your code reviews by some other engineer. Vercel built their own boat so that they can actually start reviewing pull requests and suggest changes like an engineer would. Curso has a similar behavior as well, and so it’s exactly what you said. It’s like meeting users wherever they are because that’s just great UX. You’re not telling people to learn a new pattern. You’re meeting them where they are, and not just where they are, like in the way they actually used to work. That feels very natural. There’s nothing new to learn. It just works, basically, so yeah, especially when there’s a paradigm shift like this with AI, I think that’s quite important nailing that UX and making it super easy to adopt that. That’s the key to success with AI.
Turner Novak:
Yeah. One interesting thing I feel like, I don’t know if I agree with this, but a lot of people are like the UI of software is going to change because of AI. It’s all going to be a chatbot or it’s all going to be like fluid or something like that, but it sounds like almost what you’re saying is actually the way we interact with this is actually going to continue to meet us where we’re at almost. I don’t know if that’s like a good assessment on what you just said or maybe somewhere in the middle.
Steven Fabre:
Well, I don’t know, to be honest. I can’t really predict the future, but I don’t know how many times I’ve heard people tell me like, “Oh, UI is going to change, you won’t need an interface anymore. We’ll just talk to robots and then we’ll be done.”
Turner Novak:
Yeah. Like Neuralink will be like directly into our brains with like the LLMs. Who knows?
Steven Fabre:
Maybe. I don’t know. It’s hard to tell. Maybe things are going to evolve in such a way that... I don’t know. Maybe AI is going to get so smart that they’re going to guess everything that we want to do, but I don’t know. To me, I see it more like the way we’ve been working with people and collaborating with people as humans, to me, that’s the right thing we want to replicate, I think, with AI to make it easy to adopt for people. We know how to talk to another human. We know how to do that conversation. We know how to ask something from another person.
Training AI like another entity that can do work and collaborate with you is what I see working. Everything else that people have tried that I can think of, I don’t know, hasn’t yield the results that people expected, I would say. I think Chat is a great way. I use ChatGPT all the time, don’t get me wrong, and sometimes with ChatGPT, I use voice, for instance. Typically, I don’t really use voice that much.
Turner Novak:
Yeah. Well, because there’s this dynamic with voice where I feel like if you send someone a voice message, you are saying that your time is more valuable than theirs. I might have like a 30-second thing to just say and I just send it to you, and if I would have typed it, it might have taken me like a minute or two minutes to like-
Steven Fabre:
That’s so true.
Turner Novak:
... type it specifically. I might have to edit it just to make it concise, and it would take you like two seconds to read it. There’s like a hundred words or like 200 words. You can just read it two, three, four seconds, but with voice it’s like, and I might have been like, “Aah, actually, never mind. I actually changed this. Never mind. The first thing I said, disregard that.” It’s like a two-minute message and you’re just like it took you two minutes to hear it. I’m saying that my time is way more valuable than yours, which is maybe true in a lot of cases, but also with AI, you can just be like, “Listen to this.”
Steven Fabre:
You use Voicenotes, Turner?
Turner Novak:
Sort of, not a ton. It’s usually if I want to send someone a pretty long message and I’m just like, “This is going to take me like 30 minutes to polish this thing-
Steven Fabre:
It should have been a call.
Turner Novak:
... “and I’ll just send you three minutes.” Yeah, it’s kind of like a weird in between of call. Should this be a text? Should this be a meeting? Should this be an email type of thing? It’s big in Europe, isn’t it? Don’t Europeans generally send a lot of voice messages?
Steven Fabre:
Yeah. I moved back to Europe. I mean, I live in Paris now after living overseas for 10 years. I was in New York before and Sydney and I’m shocked. Everybody’s sending voice notes here. I’m starting to send voice notes myself because that’s what you end up doing, but it’s helpful. I think sometimes it’s easy, but you’re right. It’s like I’m too lazy. I’m just going to record myself. It’s going to be easier for me, but on your side, maybe it’s going to take you longer to process whatever I’m saying, but yeah, I feel like AI is actually pretty good at... When you’re talking about something quite complex, I think voice is quite good because you can go in sort of rabbit holes. AI is actually pretty good at resonating about that stuff versus if you had to think about how you’re going to structure the prompts in a way that sort of makes sense, that it’s going to take you a long time for sure, so...
Turner Novak:
Well, and it’s kind of AI is good at that. You think of the note-taking software or whatever. If we were recording this entire podcast, it’s like an hour long, whatever, and at the end, it gives us a summary that takes us five seconds to read a couple of bullet points. The best way to incorporate AI and software is collaboration. Steven thinks that design is very important. He mentions at the end that he likes basketball and they talk about basketball or whatever we end up talking about here. You’re like, “Cool, I got it. AI summarized it. I did a good job. I don’t need to actually listen to this thing.” Yeah, that’s the superpower of this stuff.
Steven Fabre:
For sure, but yeah, I don’t know exactly what’s going to happen. I’m a bit hesitant thinking that everything’s going to change. Sure, I think there’s going to be some paradigm shifts. What I think is likely to happen is that a lot of the stuff we do in Google today or what we search, probably a lot of that stuff’s going to happen in ChatGPT or Cloud. I think a lot of the browsing experience might even happen in the chat in some extent. You already see this, but if you want to purchase some products, having those products kind of being surfaced as a component within the chat that you can quickly see the item, go through the photos of the item, click Buy It Now. I think I can really see that feature because you’re in that case providing a better experience and a lot of people have already learned to kind of go to that chat interface.
It’s sort of like meeting people wherever they are now. It’s still following this principle that we just talked about. I do see some changes happening there, and so a lot of companies will have to think about how to be visible in those AI conversations inside of ChatGPT or tools like Cloud. There’s something there for sure, but I think this is more like an SEO problem more than a how-to-work-with-AI problem, I think.
Turner Novak:
Yeah. Well, there’s been kind of like this massive surge of sort of GEO startups. It’s like SEO. For people who don’t know, it’s like when you check ChatGPT, you ask it a question, it’s like the act of showing up in there. That is a huge thing that people are trying to solve now, and there’s a dozen of these startups that are, I think at the time, currently making a decent amount of money just helping companies show up in the search results.
I don’t know. If you look at historically how those types of companies have performed, generally like marketing technology, they’re hard businesses because it’s hard to differentiate and you don’t really own the customer at the end of the day. Google and Facebook won online marketing. All these different tools that help you do marketing, it’s like people spend majority of their time on Facebook and Google and they run the ad network.
Steven Fabre:
They know the algorithm is, I guess, what you’re saying? They sort of control what people end up seeing, and so is that what you’re saying? Then, the company’s building on top of that, they don’t have the full control and so...
Turner Novak:
Well, yeah, and Google can just be like, “Yeah, we’ll give you these things for free because we own the real estate and you’re paying us for the ads anyways and, oh, we’re just going to increase the prices 20% this quarter and you just have to pay them anyways because if you want to reach people, we control the internet and you won’t reach these people if you don’t pay.” I mean, generally with like an internet based business, whoever controls the distribution to like the end customer or the end consumer captures most of the value at the end of the day. I think it’s probably going to be the same in AI. It’s just like a question of how much it shifts. Does ChatGPT cause Google to go to zero? Probably not, but does Google’s market share go from 95% two years ago down to like 40%? That’s possible. Yeah.
Steven Fabre:
Possible. Yeah. Yeah, that’s super interesting. I’ve tried some of those tools and it’s quite mind-blowing the shift that has happened. I think for us, I believe ChatGPT is probably third or fourth in terms of, or maybe fifth, I don’t know, but it’s in the top five for sure in terms of like the traffic that we get on our website. It plays a pretty critical part, like people asking questions on how to build a commenting system or like a collaborative app like Figma. It’s super important for us that we show up in those results because, and like Google, people can actually get actual get-started steps right from the chat. It’s pretty big adoption and distribution funnel, I think, for moving forward, so-
Turner Novak:
Yeah, because I think I’m like kind of a boomer in the sense, but I sort of like with Google you get all the links and you can like kind of tell what you’re going to get from them versus with ChatGPT like-
Steven Fabre:
Sort of hallucinates like this in some ways.
Turner Novak:
... yeah, but it also removes choice. It’s just like, “This is it, this is what you must do,” versus with the 10 links you get on Google, you can kind of see like, “Oh, this is from like this one website that I’ve used before and I know I trust it. This one looks good from the preview. This one, it’s like a Russian URL, so I don’t trust it or something.” You know what I mean? There’s different, I guess, like a little bit more of metadata around the information that you’re getting with Google.
Steven Fabre:
I see you’re getting more context before you actually click on the thing. Essentially, I never thought about it that way, but that makes sense. Yeah. Well, yeah, with ChatGPT, which is kind of the main LLM that I use personally, I’m always cautious about like, “Ooh, I think this is directionally right,” but like I don’t have this... Sometimes it does happen to me quite a bit, to be honest. It does tell you things that are not true or like even when you ask it for the source, sometimes it will tell you the thing, but like the source is actually not a real source or whatever, so-
Turner Novak:
Yeah, and one of the things that’s kind of been brought to light lately is that majority of this ChatGPT sources is literally from Reddit, and who is putting it on Reddit? It’s the company saying, “What’s the best search engine? Oh, Google. Google’s great. You should go to Google.” Or like, “What is the best food if you want to be really healthy? McDonald’s has X amount of nutrients.”
Steven Fabre:
Reddit plays a big part in training the models. I started exploring a lot like the... I forgot how you call it, the AI SEO.
Turner Novak:
GEO.
Steven Fabre:
GEO, yeah. One of the best practice thing you can do is actually post on Reddit and post it in a way that doesn’t feel spammy, something that feels really genuine because Reddit is actually pretty strict about that stuff. That’s how AI gets trained and that’s how you show up in AI results if a lot of people are giving your company or your solution as the solution for the specific topic people are asking for. It’s quite... yeah, it’s interesting.
Turner Novak:
Yeah. I mean, it’s like one of those things you do want to be genuine and honest as the company and also as a consumer. You should be like... you should just understand. Personally, I think it makes it harder to vet stuff because it’s hard to know who controls the Reddit account. I don’t know if I should mention this on the podcast, like the whole Epstein files thing with Ghislaine Maxwell, his secondhand woman, she was like a top Reddit moderator on multiple subreddits. Did you know this?
Steven Fabre:
I had no idea. For real?
Turner Novak:
Yeah, yeah. It hasn’t been specifically proven, but there’s been the account that they thought she was linked to, it stopped posting the day she got arrested and taken into custody. And it was a top moderator and top poster, that basically gatekept Reddit.
Steven Fabre:
For what kind of topics though? Or stuff related to...
Turner Novak:
Like world news, shaping the narrative of how people think about and discuss just cultural and high level political things. And so I think Reddit is great. I use it all the time. For sports, I love getting information and seeing people discuss things, but you do need to understand a layer deeper of where your information’s coming from. And it can kind of be masked and the sources can be muddied a little bit. Anyways.
Steven Fabre:
I don’t use Reddit much, I’m more of a Twitter guy, but it has pros and cons as well on Twitter.
Turner Novak:
Yeah, Twitter’s another one. You just got to know.
Steven Fabre:
Yeah, you’re pretty big on Twitter, Turner. I feel like you’re the meme connoisseur of VC meme.
Turner Novak:
Yeah, it’s a good strategy. Memes are a good marketing strategy that a lot of ... I think people know that now.
Steven Fabre:
It works. Maybe we should do this for Liveblocks, I think, do a bunch of memes around... I don’t know what the topic would be, but yeah, it definitely worked for you.
Turner Novak:
Yeah, one thing you’ve done, you’ve done a really good job just generally marketing, and I think launching products, generally speaking. You get a lot of attention even though you’re not really doing memes. I think you’ve gotten to number one on Product Hunt multiple times.
Steven Fabre:
I think we got number one once, I think, and we got multiples were we’re in the top three or top five.
Turner Novak:
So what’s a general framework for doing well with a launch?
Steven Fabre:
Yeah, I have a lot of thought about this. I think Product Hunt, it depends on your market. I’ve had launches where I was like, “I think this was a lot of investment for not a lot of returns,” but I also have had the opposite. So I think in our case, what’s challenging with Product Hunt is that we are a dev tools company, and so developers are not necessarily on Product Hunt.
Turner Novak:
I feel like they’re more Hacker News.
Steven Fabre:
Yeah, that’s more Hacker News, or Twitter, or Reddit even. And also developers are allergic to marketing, I think, and marketing BS, they can really smell that. But still for us, I think it’s important to be present with product people, which tend to hang out on Product Hunt, and so having some visibility there has been super helpful.
So I think in terms of being successful on Product Hunt, there’s a lot of things you need to do. You need to prepare a lot in advance. So you basically want to prepare what you’re going to post on Product Hunt, like beautiful images that you’re going to post there, a really nice demo video that’s really well-designed and put together.
Prepare the right content. And you want to also prepare things that will create some type of buzz, especially at the beginning of the day, because people who are listening... The thing with Product Hunt is that it’s basically happening at midnight, Pacific Time. So as soon as it’s midnight, that’s the beginning of the launch and it lasts for 24 hours. And so the first four to six hours are actually pretty critical because that’s when Asia is starting to vote and then you get into Europe. And so you want to make sure that that initial traction is there because otherwise you’re not going to be on the top of the list.
Turner Novak:
So you need to get whoever is awake at 12:00 AM Pacific. So that’s India. Is that India? It’d be lunchtime for them.
Steven Fabre:
Usually what I do, because for me, because I’m in Paris now, it’s at 9:00 AM for us. So at 9:00 AM you typically get Europe. You also get evening time from Australia. So basically what I do, every time we launch something, I have a list of investors or customers that really like Liveblocks, people from our community, and ahead of time I actually tell them. You probably got some of those emails Turner yourself.
Turner Novak:
Yeah. Yeah, I get them all the time.
Steven Fabre:
I’m sorry if I’ve been spamming you for this.
Turner Novak:
Do you set those up with an auto sender or are you copy and pasting from a good Google Doc?
Steven Fabre:
A lot of copy pasting actually. I tried to automate it at some point. I tried to automate some of that stuff, but it didn’t really work because I think when you automate things and you send actual... Then it turns into like an email newsletter, and so then your email is going to be in the promotion folder.
Turner Novak:
The spam folder or whatever. Yeah.
Steven Fabre:
Yeah, or spam folder or something like that. So what I’ve done is either you BCC people or you DM them one by one. So what I’ve done in the last launch, and that’s where we finished first, I actually DM people one by one, wherever they are. WhatsApp, SMS, Slack, email, and I keep track and I literally do that with hundreds of people. It takes me an entire day to do it manually.
Turner Novak:
Yeah, so you have to figure out, is it worth being number one? Do you get enough to spend that entire day?
Steven Fabre:
The visibility you get from that is definitely there. You definitely see an uptick in terms of visitors and awareness. So it really helps.
It depends, I think, what your space is in. I think for a dev tools company, it matters probably less, to be honest, because the end user is the people that actually uses the product and not the people that are on Product Hunt. But for us, the people that are on Product Hunt, they’re adjacent to Liveblocks. And so they’re actually pretty likely to recommend us to engineers if you’re a product person or a designer. If you see that what we push out is actually well-designed and well put together, I’m pretty convinced that people would then talk about it. And so yeah, I’ll probably still do it, but it’s definitely a lot of work and investment.
There’s another benefit I would say around Product Hunt, it’s not just like the amount of inbound that you get from the people seeing that on Product Hunt or social. I find it good practice to have moments and marketing moments internally, because that creates a bit of urgency and momentum internally within the team to actually complete something by a specific date. And even if that is a little bit artificial, it actually enables you to move faster and also like have something worth celebrating as a team once it’s done.
So I think about those things in terms of it’s kind of like a heartbeat. You have big launches, big marketing moments, and then maybe you have smaller updates throughout the year. For us, I always try to have like two, maybe three max, what I call big marketing moments, which turn out to be Product Hunt launches, or big version updates, or launch weeks, those tend to be big moments. But then you want to also have, like in between those moments, you want to make sure you keep showing momentum internally. And so having blog posts and updates and things like that is the framework I’m using internally.
So I guess this was a bit of a rabbit hole, but I hope that answers the question.
Turner Novak:
I feel like you have a couple... I think you’re on Liveblocks 3.0. I think. So 1.0, 2.0, 3.0, how’s the product evolved over time?
Steven Fabre:
Yeah, so we started the company in March 2021, so we’re getting close to five years now. Crazy, looking back. First version, so Liveblocks 1.0, chapter one was very much focused on building a real time infrastructure for developers. And what we offered was fairly low level APIs to build real time applications like Figma.
We had pretty good traction with this, but what we’ve learned is that if you wanted to build a Figma-like application, or Google Docs-like application on top of that infrastructure and those low level APIs that we provided, there was still a significant amount of engineering work to do.
Turner Novak:
Like creating like the actual interface?
Steven Fabre:
Basically creating the interface and all the patterns and UX on top of that infrastructure. The infrastructure was actually just a tiny part of what made those products great.
And so that was a big learning for us because while we saw a significant amount of adoption, that enabled us to raise more money and continue to grow the company, we actually struggled to make this a decent business. And I think there was just way too much engineering work on top of this to really build an amazing collaborative product like Figma.
So then we entered chapter two, and that’s Liveblocks 2.0 and that’s when we started shipping pre-built components for specific collaboration use cases. So we started looking at different verticals and like, “Okay, what makes a great collaborative software like Figma, Notion, Linear, Google Docs, et cetera?” And what we find is that across all those products, they all had commenting system, they all had notifications, and a majority of them had a collaborative text editor piece in them.
So what we started doing is, “Okay, let’s actually build pre-built react components for comments. Let’s do the same thing for notification inbox. Let’s do the same thing for a rich text editor that feels like Google Docs.” Then when we started shipping those pre-built components, we’ve seen like a massive inflection in terms of our growth because now those components enable developers to access all the infrastructure that we have built, but the time-to-value was just like shortened by a significant amount.
And so if people were looking for like a specific use case, “Okay, I need a commenting system in my app with notifications, so that when people are mentioned, they get notified.” Or, “Actually I need something that feels like Google Docs or Notion, that’s like a text editor when I select text, getting all the notations in context. I can do that kind of stuff.”
Now we sort of narrowed our ICP, but those people that needed this, it was much, much easier to adopt Liveblocks. And so that was the second chapter for us.
Turner Novak:
So it was like all the things that people were going to use the infrastructure for, you just built it for them so they could super easily roll out all the features instead of having to build it themselves.
Steven Fabre:
Basically, instead of just giving a low level infrastructure, what we started doing is actually providing the interface on top of that infrastructure, which is a bit more opinionated. You can still customize a lot of that stuff, but it’s opinionated and it does one thing really well. That really helped.
And so now we’re entering chapter three, version 3.0 of Liveblocks, and this one is very much about leveraging all the stuff we’ve built to enable people to work together, but bringing AI into that story so that people can now actually collaborate with AI using very similar patterns, as I said earlier. Being able to mention and AI inside of comment, being able to mention an AI inside of a text document. When the AI is actually doing work, being able to be notified when that work is done, wherever you are. You’re in your email, or in Slack, or whatever the case might be. So meeting people where they are with building that infrastructure and all the orchestration for that.
Turner Novak:
Is that really that hard? Why do you need Liveblocks for it? It’s like an ad mention for a thing. I don’t know, what do you guys solve that’s so hard for developers?
Steven Fabre:
Well, I think what’s really hard there is the UX and making sure all the pieces work well together. Sure, for comments, for instance, you could build a pretty basic commenting system yourself, but when you think about enabling things like attachments, emoji reactions, mentions, making sure that when somebody is mentioned, the system is actually aware if that person has read that comment or not. So that when you notify that person, you actually don’t spam them, you only send them when they need to actually see the notification. There’s a lot of logic and things that make an experience feel really great.
So when you start to add those things across comments, a collaborative experience where people can actually edit a document together simultaneously, there’s a lot of complexity in those areas that basically using a tool like Liveblocks saves you a ton of time. First, when you go-to-market, but also the maintenance aspect of things after the fact.
So yeah, our focus is very much on making sure the orchestration is handled and works beautifully across the board. And I would say our expertise as well is around the end user experience. So we spend a lot of our time looking at some of the world-class tools out there like Google Docs, Notion, Figma, trying to reverse engineer some of the patterns and things that they build, and try to make that available for their companies so that they don’t have to do that themselves and spend a ton of months. It could take months or years doing that yourself in house.
Turner Novak:
So it’s Q4 of 2025 right now, I know back in 2024, you started to get pulled towards more of like enterprise customers, larger, more established products that were starting to use you. I remember when we first met back, you were like, “Oh, new startups are building on Liveblocks.” And now you’ve evolved into, there’s still that, but now there’s also, I think there’s a company that’s 40 years old that’s a customer. Right? So what was happening over that time?
Steven Fabre:
Yeah, definitely. I think for the last year and a half we’ve definitely seen a shift. Going from indie developers and smaller startups, to bigger companies starting using Liveblocks. That was a huge learning experience for me because I’ve never really done enterprise sales myself. So yeah, really had to focus on that, learn a ton, still very much learning. But yeah, we definitely made that shift and so that enabled us... Basically one thing we wanted to prove is like, “Okay, is there a meaningful business behind Liveblocks? Can we actually sell to enterprise?” And we’ve definitely checked that mark now.
So yeah, I took an advisor on the go to-market side, worked with them for six to nine months, really helping me grow the revenue to, I think at the time, about a half a million dollar, which is not that much looking back.
Turner Novak:
Significantly larger now?
Steven Fabre:
Definitely larger, yes. We’ve grown quite a bit.
Turner Novak:
Well, I think a couple customers you maybe mentioned, I don’t know who you’re allowed to say publicly, but I think I’m allowed to say them. You have people like Cisco, PWC, Rippling. I think Magic Patterns is a younger starter that people probably heard of. Typeform I think uses it. Resend.
I don’t know who you’re allowed to link up with the product, maybe they can get mad at me for saying their names, but what are maybe some of those companies or some other ones, what are some things that they’ve done that people might be interested in?
Steven Fabre:
Yeah, for sure. So a good example, Rippling, you mentioned, they’ve recently released a one-on-one tool. So basically Rippling is a HR platform, HR tech, and so they have companies using their tools so that their managers can handle one-on-one conversations with their direct reports. They built this tool with Liveblocks. And so they have a Google Docs-like text editor that’s real time and multiplayer. So multiple people can edit it simultaneously, and they can leave annotations and comments right there, and mention the people so that they get notifications. And they built all that stuff with Liveblocks.
Magic Patterns, they recently raised the Series A. They built their entire product foundation on top of Liveblocks. So they have, for those who don’t know, Magic Patterns is like...
Turner Novak:
Like a AI designing tool?
Steven Fabre:
AI version of Figma, I suppose. Like it’s a visual creative tool to build landing pages and apps visually on a canvas. They’ve built this with Liveblocks. So it’s a canvas where you can collaborate, you can see live cursors with people when you’re together in the same file simultaneously.
They also integrated comments and notifications from Liveblocks in there. So if you have a specific feedback on a part of a design or landing page that you put together yourself, you can quickly add Turner, like, “Hey, can you change the color over here?” And then you’ll be getting a notification and you’ll be able to make that change on your side.
So those are just similar examples and you can see that the breadth of the kind of customers we have from early stage startups growing, and also larger established companies like Rippling.
Turner Novak:
So one thing you mentioned I thought was interesting, we both kind of skipped over it, but you’re learning how to do enterprise sales. I think that’s a lot of things, a lot of founders, average founders, probably somebody who’s like, “I don’t want to do enterprise sales.” It kind of sounds dirty, it sounds hard, it sounds like it’s hard to do. So have you evolved as a founder in terms of understanding how to do that?
Steven Fabre:
It’s definitely difficult for me because I’m a designer and I love making stuff. And so learning how to actually make money as a business and selling is-
Turner Novak:
Yeah, a big deal. It’s the point of all this.
Steven Fabre:
... pretty important as it turns out. But I think honestly, I enjoyed it. What I love personally, as long as I’m learning something new, that’s enjoyable to me.
Turner Novak:
Is there a general thing that you’ve got, like a framework you picked up on is, like, “Oh, this specifically really helped me learn how to sell a product to somebody”?
Steven Fabre:
For me, I don’t know if I have a framework, I’m not a big framework kind of person, but I know myself enough that if I know I need to do something and learn something, the best way I learn is actually by doing it myself. I need to do it. And so I try to put stuff in place to sort of force me to do the things, especially if those things are not things that would come naturally to me.
And so in the case of like go-to-market and sales, to really prioritize it and treat this seriously, what I decided to do was to actually work with an advisor and literally have somebody join all of my sales meetings for like two months. And then have two hours a week blocked out with that person so that we could just focus on this and talk about this. And I knew that by having a schedule in my calendar, it forced me to treat the thing seriously. Because I knew I would have this meeting weekly, I knew I’d be talking about it, and so it’s like a way for me to enforce that. I need this enforcement sort of thing, otherwise sometimes I can put things aside.
And so that’s what I did. I worked with somebody called Dakota McKenzie and he’s really helped me. Joined the sales calls, giving me feedback after the calls so that I could learn. And then with them, we went through different pricing changes as well. We understood some of the pushbacks that we were getting. Honestly, that helped me build a better product with the team, because some of the feedback I was hearing helped drive the roadmap a little bit better than before.
I think that before we were unconsciously building something the engineers on my team wanted, or something that I would feel like I would use myself. But what we found is that when you start having more conversations with customers and true enterprise potential customers, there are challenges and the things they want you to solve, the foundation is the same, but there’s subtleties and differences that is what’s going to enable you to win some of those deals.
So that’s how I’ve learned it. And I really wanted to grow that revenue myself at first before handing it off to somebody else. And so once we got to, I think 400 or 500,000 in revenue, I ended up hiring Stacy. And she’s been amazing, she’s been taking that on her plate. And yeah, I think you’ve talked to her briefly, but she’s incredible.
Turner Novak:
Yeah, she actually gave me some ideas of stuff we should hit on in the podcast. So shout out to Stacy, if she’s still listening right now. I think she’s going to listen to the whole thing. So we have at least one listener that’s still ...
Steven Fabre:
Hopefully more. Hopefully more. But I’ll text her, “Stacey, if you don’t listen to that point, I’ll know.”
Turner Novak:
Slack, Steven, right now, when you’re listening. Just issue the message.
So it sounds like one of the things you learned from those two hour structures, like, “Let’s go back,” is almost like listening to the customers and what they wanted versus what you wanted. Was there anything else that you learned that you changed throughout this process? Like how you evolved in your go-to-market motion or sales motion?
Steven Fabre:
Yeah, definitely. And I’m still very much learning. And I think I was giving away too much for a while. I don’t know, I think I’m a nice person. I don’t know. But I feel like because I hadn’t done really sales before, my natural default was like, “Yeah, sure, we can do this. “ I was trying to be really helpful and I felt that I was always going to come back. That’s true in some ways, but I do think you need to learn how to draw the line. And so understanding that dynamic of like where you draw the line and making sure people see values, like I give you something, you give me something. Nailing that dynamic is something that it’s more like an art than a science, that I feel like I picked up.
But at the same time, I would say that I’ve made some mistakes, to be honest. And that’s something I’m realizing now, like a year and a half later, maybe two years later at this point, is that because we focus so much on going towards enterprise, we’ve made some pricing changes and changing to the business model along the way that I think impacted what made us successful in the first place.
And so as we went towards enterprise, we made pricing changes and things that were more optimized to convert bigger customers. And I do feel because of this, there are probably some early stage companies or indie developers that maybe didn’t even consider Liveblocks. Because perhaps they saw us more as like something for enterprise, or maybe pricing wasn’t as transparent as it should be. And in retrospect, I don’t think that’s necessarily a mistake because we needed to do that to grow at the pace we needed to grow at. Yeah, I’d probably do things slightly differently if I were to do them again. I don’t know if that makes sense.
Turner Novak:
On pricing, is it just like there’s too much of a minimum check size barrier or something? Like you can’t use it for free? Or if you’re going to pay, you got to pay at least 10 grand? That kind of thing.
Steven Fabre:
Yeah, I think a pricing model became a little bit too pricey for some type of potential customers. And so yeah, I would say that’s the big point. That’s the major factor there that impacted us. But that’s also what enabled us to increase the ACV on the enterprise side.
And so now what we’re doing is trying to shift, but again, so that we can have a really solid self-serve without losing on the enterprise side of things. And so that’s a big project we’re working on right now to improve pricing. Once again, I think pricing is never done, but I feel like this is something I’m quite excited about. It’s really taking Liveblocks to our roots and making sure anyone can actually use this, whether you are a teenager messing around and building fun little side project, or if you are a startup or a big enterprise, we want to make sure that Liveblocks can work for you. We want to make sure we have a business as well on our end so that we don’t end up giving too much away for free. So, that’s something we’re working on right now. But yeah, I’m quite excited about this, especially in this day and age. And I’m not sure... You’re probably seeing this, Turner, as an investor, but the expectation now is kind of crazy. Like the growth that investors expect, I would say, when you’re on the VC-backed track. You see companies growing from zero to tens of millions in a year, or in weeks literally, a couple of months. I think the expectations when you have pretty solid growth, what would have been world-class 3, 4 years ago, now it looks average. The triple, triple, double, double, whatever. That’s kind of average now.
And so I think it’s quite important to think about this, because I think historically the playbook was probably, first, you build great product that people love. Maybe it’s going to be for startups at first and SMBs. You probably grow the revenue to a million or two million in there. And from there, you’re triple, triple, double, double, double, and that gets you to 100 million. And the way you would typically do this is by scaling things up. And scaling things up means you need to get into enterprise once you find product market fit, and getting into enterprise means hiring a bunch of salespeople. People that do marketing, and people that do solution engineering, and customer success.
I feel like that has changed a little bit because that’s like quite a significant investment to do this. And with a lot of those native companies, what’s been shown is that you can get to pretty significant growth and fairly high revenue numbers without as many people. And so, that’s one thing I’m focused on right now. I want to make sure that we can have a business that scales in terms of revenue without necessarily hiring a bunch of salespeople. And that means having a really solid self-server I think is way more important now than it used to be, I would say.
But I don’t know. I’d be curious to see what will you think about this, Turner, as an investor. I don’t know if you’ve seen similar patterns or... Yeah, I don’t know how your portfolio companies are doing on that front.
Turner Novak:
Yeah, I have one portfolio company that’s probably... It’s always going up, but he’s probably like around 8, 9, 10 million in revenue run rate. I don’t know when exactly this will get published. You could probably fact-check me and I’m going to be off on this number, but it’s like roughly around that. And it’s just the founder doing sales.
The ACVs are pretty big. Some of the ACV’s six figure. He might have like a seven-figure ACV. I think that’s why the companies are scaling really quick is because you can get a seven-figure contract with someone. Someone will pay you over a million dollars to solve this problem. And it’s because, a lot of the times with these like AI native product, they’re doing some work. And it’s almost like not just software budget, but labor budget in a way, that gets woven into the product. So you are able to basically say, “Hey, existing structure, the way you do this, there’s like 10 people on the team, and this is like a $10 million budget line item. But with us, you only have to have two people on the team, and you can instead have a $5 million budget line item.” Some of it is people, some of it is software. Or, “You’re still paying $10 million budget line, but we’re driving so much more revenue to you because you’re using our product.” So, it’s completely justified. The ROI is clearly there. I think there’s some elements of that.
There’s also some elements of like, we have an AI innovation “budget” and we just need to spend it. So, we have $100 million or 50 million to just... So, we’ll spend seven million or seven figures on just AI search or whatever. Hopefully, that’s not one of the categories, but maybe it is. But there’s definitely some categories like that where it’s just like, we’ve got a budget to spend. And if you’re a Fortune 100 company and you have 50 million bucks you just spend on AI, and if you don’t spend it, it doesn’t get spent? You’ll just spend it on stuff.
Steven Fabre:
We’re already seeing this. That’s where some of the big companies are like... Some of them can spend a significant amount with Liveblocks is like... This US has this AI budget spend. And so...
Turner Novak:
That’s bad. You don’t want that, do you?
Steven Fabre:
I think there’s a lot. Every big company, especially the Fortune 500 companies, they have budget allocated for this because they’re all scared about up-and-coming startups taking their... Like, eating their lunch, I suppose. And so, yeah. I guess, you want that. Why wouldn’t you want that money?
Turner Novak:
I mean, you would want the money, but you want it to be tied to solving a problem for them and their customers.
Steven Fabre:
It needs to be solving a problem, for sure. Yes.
Okay. I see what you’re saying. So you’re differentiating from like pure R&D project, like, burning this money essentially and see what happens versus actually solving a problem.
Turner Novak:
Yeah, like d’you remember with crypto and Web3, it was like, “Oh, we’re a concert, we’re a music company, and we’re going to...” Instead of publishing on Spotify, it’s like an NFT. You listen to the NFT and stream it, and some company probably spent like a million dollars marketing or whatever. It’s like, no one cares. There’s probably like eight people that actually streamed the NFT music versus the millions on Spotify. It’s like, what?
Steven Fabre:
Yeah, I see what you mean. Okay.
Turner Novak:
I think, really, to be create enduring... You want to create enduring value for your customers. You want to be able to say like, “Hey, we’re using Liveblocks to actually help you increase your revenue or make your product better.”
Steven Fabre:
Definitely agree with this. The point I was getting at before was more like, sure, you could do this. And I think the old playbook would have been... Okay. Let’s say you want to go from 1 to 10 million in AR. The way you would do this 3, 4 years ago was by hiring, literally, 10 to 12 salespeople, assuming a one million quota per sales. And that’s how you would scale to those revenue numbers. I do feel like the expectations now is to have more efficiency as a business, and that doesn’t necessarily seem the most efficient nowadays. Sure, you still need salespeople. Depends on your business, obviously. But my take here is that while I still want to grow the go-to market team, I’d love to get more revenue from like self-serve so that we have a business that can really scale and we don’t need as many people to power that business behind the scenes.
So that’s kind of the direction I’m taking, which, I don’t know, we’ll see if that’s the right direction. But I feel good about it. We’ll see how we go.
Turner Novak:
One interesting direction we go from here, you’re a designer. We’ve talked a lot about design. You’ve actually built some design tools multiple times throughout the years. What was it like transitioning from being a designer to being a CEO? And then just why is design so important?
Steven Fabre:
It’s been a learning journey, for sure. And I love learning, so that’s been great. I wouldn’t trade what I’m doing now for anything else. I mean, I love building. I like making stuff, basically. If I were not working or doing anything, I would probably still be making stuff. And I think humans were designed to make stuff. I think that’s probably what makes us happy, when you create something for the people to consume or... I think we’re creative creatures. And so yeah, I personally really enjoy making apps. I find a pretty sweet spot in designing and building creative tools in the past, tools that people use to actually create something themselves. And so, I found that quite interesting from a design standpoint because those tools are quite... They’re super interactive. And it’s not like there’s like a single flow of like, “You go from one page to this page to this page.” There’s a few different options, and that’s about it.
When you build a creative tool, there’s pretty much an infinity of options in states, and that makes designing those experiences quite interesting. And so yeah, I got into Liveblocks by accident, I would say. Or not by accident, but it’s basically solving a problem that I faced myself while working at Envision. Figma was starting to become very popular, and I ended up leading design at Envision to convert inVision Studio, which was a competitor to Figma, from a desktop-based, file-based application to a browser-based app.
And that’s where I learned that building collaborative apps that were super interactive, real time, multiplayer, that was really, really complicated. And to me, it was obvious that, pretty much, most software moving forward should behave that way, because the browser is just the best distribution channel and it should just be all there. And when you want to collaborate with someone, you just share URL, and you’re together in the same space.
Turner Novak:
Because you can just be like, “I made this thing in the software. Here’s a link that I’m sending you,” and you just click it and open it.
Steven Fabre:
Instead of in the past, you would have to save a file, send the file to somebody by email. Sending the file can take time to upload and then for the person to download.
Turner Novak:
By the way, you got to download the software too, yeah.
Steven Fabre:
Yes. Yes, exactly. And then you have this artifact that you downloaded from like an email, you want to give feedback about it, change it, modify it, then you have this sort of dead-end entity that you created. You share it again, and it creates all kinds of dead ends, if you think about all the different things that... You have this thing that initiated here, and then a bunch of different branches have emerged from this, like in different areas, and training to reconcile that stuff. It’s just terrible. It’s just a terrible experience for people to collaborate that way. And so that’s why I think tools like Google Docs, Figma, and browser-based tools in general have been really successful. They solve a lot of that back and forth.
And so to me, that’s how I ended up starting Liveblocks. I felt like this was a big problem. Thankfully, I had met my co-founder at the time, Guillaume. Amazing engineer that I worked super closely with at Envision.
Turner Novak:
So, what’s the story about meeting him? I know it was a pretty... I mean, you’ve told me the story before, but how’d you guys meet?
Steven Fabre:
Yeah, that’s funny. I had been at Envision for about four years already at that point when he joined Envision, and his first day was actually at an offsite that we had in San Francisco. Spent a day together, and eventually we figured... I think I was the only French person at the company, and Guillaume was also French. I was like pretty picked that up pretty quickly from when I met him. But what was funny is he asked me once, “So, where are you from? Where did you grow up?”
And I was like, “Oh, you wouldn’t know. It’s a small town. I grew up in the southwest of France.”
He was like, “Oh, wow. Me too.”
I was like, “But which town?”
“Oh, you wouldn’t know. It’s close to Toulouse.” Which is the biggest main city there.
“Oh, me too. I was from there.” Then we narrowed it down and we’re basically from the exact same region, which is a super rural region of France called Avignon. And it was funny, just the fact that we came from a super tiny area of France. It’s pretty unlikely that you meet in San Francisco like this and we’ll be on the same team, and eventually go on and start a company together. So it makes them a good story, I think. And yeah, we just love working together after this. We often paired up on projects. And so because I love creative tools and that’s kind of my sweet spot, we ended up building or started building a presentation/video tool. And we wanted to make a mix of Apple keynote meets video tool, and make that a browser-based experience like Figma.
Turner Novak:
So it was like slide editor with video native kind of?
Steven Fabre:
Kind of like Pitch, I suppose. Yeah, that’s pretty much what it was. And the idea was to facilitate people, make it easy for people to create those kind of interactive presentations, especially in a remote context. Which, at the time, felt like the thing that was going to stick around, especially with the pandemic that was going on at the time.
But anyway, six months into that project, we realized we were spending all of our time building the real time and the collaboration infrastructure for that product instead of the actual tool itself. And that’s when it clicked. I was like, “Dang. We had a full team working on this for a year and a half at Envision, and we couldn’t pull it off.” We pulled it off, but the initial version was average. It didn’t work exactly like we wanted. It was slow and buggy.
And now here we are, just the two of us, super focused. And we’re spending all of our time trying to figure this out again.
Turner Novak:
Yeah, from scratch.
Steven Fabre:
Exactly. And so we’re like, “Damn. What if instead of building this presentation tool, we’re actually building this infrastructure that would enable the companies to build Figma-like products?” We would have needed that at Envision. We would have needed that as we started this company. So we thought, “Yeah, why not? Let’s go for this.” That’s how we ended up starting Liveblocks and ended up being CEO, obviously. I think CEO can mean a lot of different things, but I would say at an early-stage company, it’s probably very different than when you’ve started scaling. Right? I think where we’re at right now, we’re about 14 people on the team.
Yeah, I see myself as the person responsible for really setting the vision, making sure everybody’s aligned, making sure the moral on the team is... Making sure we have momentum. Those are the things I’m focusing on. And I think there’s a lot of similarities, I would say, between design and running a company in a way. It’s like... I don’t know. I think about design, at its core, is problem solving. And I think running a company or doing any kind of project in life is about solving problems as well. So, how do you design a solution that solves the problem you’re facing? Whether that problem could be, “To build X, we need X amount of capital, some amount of people. How are we going to actually solve this?”
People are not really... The vision is not very clear for people. Like, “How can I solve this? Maybe make it clear by my communication style, the way we communicate about it, or having some frameworks internally to make sure this happens.” So, designing the engine that ends up... A company is basically a tool. You’re designing a tool that outputs something else. And so, I find there’s a lot of similarities. It’s a bit more abstract, I guess, when you’re running a company, but it’s definitely interesting.
Turner Novak:
So design is not like how something looks, or does it have pretty edges or something like that? It’s more so just like how does a problem get solved for the person that has the problem?
Steven Fabre:
It’s definitely how it looks, for sure, but it’s mostly how it works and how it feels and the problems it solves. A given problem can be solved in many different ways, and design is solved the one solution to that problem. And so, designing a company is...
Every company is different. Every company is designed in a slightly different way. And a lot of those decisions come from the founder, and probably the first designers that you have on your team. They really shape what the company is going to look like.
Turner Novak:
What’d you learn, or how did you learn how to be a CEO? I know you’ve had some side projects over the years. I don’t know if you learned things there. Or was it just like a slap yourself in the face of like, “Man, I got a team I got to run”? What did you learn?
Steven Fabre:
I’m not sure. Always, even when I was 15, I think I had side businesses. I was selling websites to people. I always liked... I like making money, to be honest, from a young age. I liked making stuff for people and I always had a business mind, I would say. So looking back, I felt like I always knew I wanted to have my own business. I started multiple side projects and smaller companies in the past, so I always has this thing in me. I still like learning to be a CEO. I don’t know, because there were still 14 people, so I think it’s very different being a CEO of the 14 people company versus hundreds of people or thousands of people. I try to learn a lot from the community, and I definitely have some founders I’m really close with that are maybe a year, two years ahead of me, or some of them are years ahead of me because they have much bigger teams.
Just to get their perspective when there’s always new challenges and things that may arise, that it’s pretty likely another founder has gone through this. So I found that having a founder community has been probably the most helpful, I would say. Because those people, some of them have been going through similar challenges that nobody else really can relate to, I find. And that’s something that has been difficult, actually, starting a company. Most people don’t necessarily understand what you can be going through. And sometimes, things can be going, probably... Not that you have to pretend, but you... Not pretend that you can... I think it’s good to show...
I actually think it’s healthy to show some type of weaknesses, that actually builds stress with your team. But sometimes you always have to be enthusiastic and like, “Yeah, we’re going to do this.” I think you need to have this mentality, and that takes a specific kind of character, I would say. But yeah, having founders around has been insanely helpful. For instance, I’m working... For me, Jonathan from Maze, the CEO of Maze has been an amazing mentor to me. Their team is probably 150 people now, and they’ve grown quite a bit. They raised a Series B a couple of years ago and they’ve gone through many different challenges. And talking to him has been particularly helpful because he’s also a designer originally. And so, having him has been particularly helpful for me, yeah.
Turner Novak:
So, it’s probably most helpful to get somebody who you have some kind of a commonality with that’s a couple of years ahead of you in the journey.
Steven Fabre:
I think you can learn from all kinds of people, but I think having somebody that has the same... Somebody that cares about similar things to me makes me more relatable, I would say. And I’ve also found friends that were very, very sales-focused, and I still learn a lot from them because they have a different mindset. They don’t think exactly the way I think, but there’s still a lot to learn there. That’s why I’m trying to apply some of those things that I’m learning. I always try to be true to myself. I realize, I don’t know, sometimes it’s easy to get caught and be like, if I’m a CEO, I need to act like this. I need to do things in a specific way because that’s what I imagine a CEO should be doing.
I don’t know. The more I grew up, the more I’m like, “You know what? I’m just going to be myself, and I’m going to do things the way I think is right.” And that tends to work out better for me generally speaking.
Turner Novak:
It’s funny, a lot of people think of CEOs, you’re in a suit, in an office in downtown New York and a skyscraper, but...
Steven Fabre:
Right. That’s what I thought it was in the past 10 years ago.
Turner Novak:
You’re just on a laptop at a restaurant, in between meeting a couple random different people. You’re trying to hire someone and you’re also in Slack, commenting on some feedback to a team member who’s working on something. Or it’s like a Sunday morning, you’re in your underwear, sending something. It’s the least fancy possible thing you could imagine.
Steven Fabre:
It is the least fancy thing so far. And yeah, you basically... All the biggest programs, they end up bubbling up to you, I would say. I don’t want to overdramatize it because for me, it’s been mostly fun. But I think I’ve been mostly enjoying it. I really like it. Scaling the company, designing it the way I want, trying to create a really good experience, hire people that I would genuinely love to work with and end up working with. And I want to make sure people have...
Basically, trying to create a place that I would have loved to work at myself. That’s really enjoyable to do. And seeing people... I feel good about seeing people having a good time, to me, or feeling proud about the work they put out. That’s something I’m quite proud about actually, yeah.
Turner Novak:
Okay. I want to ask you about that, just generally, team design. You guys have a remote distributed team. But one thing that you also... I feel it’s related to what we just talked about. There’s a lot of other founders that are probably listening to this, and you talked about quite a bit you guys had a really good relationship. He actually stepped away from the company, maybe it was about a year ago. Which, to somebody who’s not a founder and not familiar with how startups work, that might sound like a death blow. Like a founder leaves, the company is over. I think it’s a little bit more common than people might realize. And a lot of founders who’ve never gone through that before might be like, “Yeah, that does sound fucking brutal.”
Can you just explain what happened, and how you moved forward as a company?
Steven Fabre:
Yeah. That was, honestly, I think the... Probably the most difficult thing I had to go through personally. And to be honest, I think it was probably one of the hardest thing to go through for Guillaume. I can’t even compare what he was feeling, compared to what I’m feeling. But Guillaume was just... I don’t want to put words in his mouth, but I think he was not enjoying it as much, running a company. And probably, ended up doing a lot of things that he didn’t necessarily enjoy. And so, I think what he wanted to do personally versus what the company needed was getting to import where it was misaligned, even if he was actually doing a really good job. He was actually amazing at his job.
And so, that was difficult because it was his decision to leave eventually, and yeah, it was really hard to handle. There was a lot of things. Obviously, our relationship first. It’s difficult, even if we’re really close. We’re friends. We’re still friends. We’re in really good terms, which is amazing to see because a lot of founder breakups end up not in the similar spot. There’s a lot of things to handle that were quite difficult for me. Basically, when that happens, you have to think about the future of the company, and you want to... There’s an equity piece aspect of this. You’re basically negotiating. You want to make sure everybody on your cap table is happy about the outcome that you end up with. One of the founders is negotiating from them. I was negotiating for the stakeholders, for myself, for the team. There’s a lot of parameters there that are difficult to manage, communication aspect, like making sure you involve investors properly so that they’re aware of what’s going on. Yeah, I’ve learned a ton handling this, and that was definitely not the most fun thing that I had to do, but I’m really happy where we landed. And look, the company has never done as well since he left, and I don’t think that’s because he left that we’re doing better, but I think it was a good reset for the company and the projects that we went after, and I would say like the moral on the team and that actually we tried to use this to create momentum, and so, that’s basically what happened.
Turner Novak:
Yeah. And I think it took you quite a while to just find the right fit. You took over engineering yourself, where we’re kind of leading it, but you were able to finally find someone? What’s the process like of bringing someone like that in and hiring your first new outside executive for an early company like this?
Steven Fabre:
Yeah, it took some time. Actually, Matias, the new head of engineering, started a month and a half ago. So it took us nearly a year, when you think about it, between the time he started, this new person, versus when Guillaume left. Well, thankfully, our team, being remote actually, I made a conscious decision of pretty much only hiring very experienced people that most of them have worked remotely and I’ve worked with before. And because our team is very senior, and I would say the vision and where we headed is very clear. Thankfully, the team has been able to continue to ship stuff, and it’s not like we had to go in a completely new, different direction or anything like that that would have created a lot of chaos. So we’re able to move forward and have a really good year with over 3X since the departure, and the team has been able to execute really well.
So I think just having the right people in place and being super transparent about what happened has really helped. I tried to be as transparent as possible with the team and tell them things the way they are, and they really stepped up. And one thing that I realized as well that I did not necessarily expect is I think when Guillaume was there, I think he was taking a lot of stuff on his shoulders, and by him leading, like by no means at that engineering level to be able to do what Guillaume was doing. But what happened is that everybody stepped up on the team, and so even if at the beginning there were some areas where I felt like we were maybe a bit weaker than we were before, what happened is that the three most experienced people and even the rest of the engineering team, they all stepped up and had to learn new things and like really show up, and I think we actually became a stronger team as a result of that, which was self counterintuitive and not something I expected.
And then finding Matias was definitely like finding a head of engineering that really took a lot of time, especially building a dev tools company for something that’s very technical, and that requires a specific kind of expertise. Yeah, it took forever. Yeah, that was pretty tough.
Turner Novak:
What ended up working? Were you just like cold, emailing tons of people? Did you work with a recruiter? Did you work with Matias in the past?
Steven Fabre:
Didn’t work with Matias in the past. I mostly worked closely with Alicia, our head of people at Liveblocks. That’s another thing we can dive into if you want. We hired a head of people very early, even, I think we were like six or seven people at the time, which is unheard of for most companies. That was a very conscious decision. But anyway, Alicia really helped me... First of all, she helped us internally understand what are we looking for. I had an image of what I wanted in my head. Stacy, on the go-to-market side, had thoughts and opinions about what we would need. The engineers on the team had a specific idea of what they wanted. None of those things aligned.
There was some other app, right? But if you think for the perfect person, that person probably doesn’t exist. It’s like you’re looking for a unicorn. It’s like, I want an amazing, strong technical person that can do engineering work, but also want this person to be an amazing communicator and being able to drive a vision forward. But maybe I also want this person to join some sales meetings because that’s something that’s quite important for us as a company. Also, as a founder, I want to find sort of a co-founder, maybe. I don’t know. There was a lot of things that when you put it on paper, that person doesn’t really exist, or maybe it exists, but it would be very hard to find it. And so yeah, Alicia really helped us work together as a team to really write things down, what do we need, what are priorities, and make sure expectations were aligned so that we could have a good process to then find this person.
Nonetheless, it was quite difficult to find this person, and I’m really glad we landed on Matias. He has a lot of expertise in AI modeling and stuff like that. Also, has expertise working on collaborative infrastructure and real-time infrastructure. He worked at Sketch for a while. And yeah, Sketch was a competitor to Figma, still is actually, and so there was a lot of challenges that we are solving with Liveblocks that he had to go through himself as well. So yeah, but we talked to a ton of people to get there, for sure.
Turner Novak:
One thing you mentioned right there was the importance of finding somebody that slot into a remote distributed team. And we talked a little bit about that before, but what have you learned running... I don’t know. It’s a successful team. I don’t know how you like define success, but it seems like it’s worked to some extent of, like running a distributed company for multiple years that continues to have momentum. What are some best practices for people?
Steven Fabre:
Yeah. So first of all, I worked remotely for a while, even before doing this. And I worked at InVision, which was one of the pioneer in remote work. Fun fact, when I joined InVision in 2016, I was actually in Sydney, and most of my team at InVision was on the East Coast in the US, so I learned a lot of what didn’t work there. I learned a lot about what worked there, and then I tried to use some of those learnings, the good and the bad, tried to apply that to Liveblocks. So, yeah.
Turner Novak:
Maybe what’s the bad? What’s the good?
Steven Fabre:
Well, the bad, I think some of the things I learned at InVision is, there was a couple of things that I think didn’t work. Like when you work remotely, you probably want to have most of your people in a specific time zone, or at least have a decent amount of overlap, otherwise it becomes very difficult. I clearly saw that this didn’t work when I was in Sydney and the rest of the team was on the other side of the world.
Turner Novak:
Because you were literally like a 11 hour, 10 hour difference, something like that?
Steven Fabre:
If I remember, I think I tried different things. When I joined the company, I was like, “You know what? I’m going to work in the evening, and I’m going to go to bed at 3:00 AM, and I’m going to adjust my schedule.” I tried this for three weeks. I was like, “I can’t do this. I’m going to quit instantly if I keep doing this.” So I shifted my schedule to start at 5:30 AM or 6:00 AM Sydney time, and I remember 6:00 AM Sydney time was like 4:00 PM or 5:00 PM Eastern.
Turner Novak:
Okay. So it didn’t even do that much.
Steven Fabre:
Yes. So, it’s people are finishing their day in Europe, they’re gone already. So you get a bit of overlap with people on the West Coast. So, anyway, I learned that that was really difficult.
Another thing that I’ve found is that at InVision, everything lived on Slack, so we didn’t really have a central place where work would happen or things were documented. That was another thing.
And then I would say there was not, even if there was, maybe I joined, we were like 100 people, so maybe that was different, but I felt there was not like, “This is how we operate at InVision. This is our culture. This is how we do things.” There were some parts of that, but I don’t think it was truly... I don’t know. It didn’t feel like there was alignment at the people that are actually doing the work in terms of how we should work and how the values, principles, and things we should follow.
So starting Liveblocks, pretty early on, I was like, “Okay, we need to be very clear on that stuff, this is how we work.” Also, knew that early on we needed to hire somebody that would be sort of in charge, not in charge, but somebody that would really look after the culture and the way we work together. And that’s why I wanted to hire somebody there, and I ended up hiring Alicia. She’s been amazing, absolutely love working with her, and just enforcing culture and in a way that feels natural and... I don’t know. It’s really helping because we’re very intentional about the way we work together. And so, one thing that we want is that we have clear, even if we are flexible in terms of working hours, we ask people to have two to three hours of overlap of work every day.
So to do this, we typically hire only in Europe and East Coast in the US. We have one person on the West Coast, Jonathan is in Seattle, worked with him for four years at InVision, and he works different hours to accommodate for this. That’s something that I think is very important. When you’re remote, you definitely want to have some overlap every day because if you don’t, some decisions that could happen today will be made tomorrow, and that, just, if you compound that over years, that slows you down by a significant amount. So that’s thing number one.
Number two, you definitely want to spend time with people in person. So we have one offsite a year where we bring the entire company together. We just came back from Punta Cana, actually. We were there last week with the team, and we try to create a week where, obviously, we work in the morning and find what are the things that we cannot do remotely to align. We try to use this time to build connections and make sure people get to know one another on a personal and professional level because once you know the people on the personal level, it’s much easier to work with them remote, and so like there’s like a... I don’t know. Something that you learn. So that’s number two.
And then, we also encourage people. We actually have budget allocated for project teams. If they sometimes need to meet in person, this is something we encourage them to do. So those are things we do, and then our tooling, we use Linear and Notion. So we try to document things as best as we can so that we can actually onboard people more easily when they join, and there’s a ton of things like that that we try to follow.
Turner Novak:
So it sounds like a lot of it realigns to make sure that there’s people... That there’s shared times during the day that everyone’s working to just have good sync, meet your team in person as consistently and as much as you can, and just spend time getting to know each other. And then, it sounds like people that are able to work autonomous. You may have mentioned this earlier, but just...
Steven Fabre:
To be, at the end of the day, it’s probably the most important factor. The most important factor is the kind of people you hire. Not everybody’s made for remote, and in that pool of remote people, there’s probably a good amount of them that are just there to just follow the lifestyle, I would say. But there’s a pool of people in there that are really passionate people that really want to do good work with other smart people. Those are the people you want to hire.
But at the end of the day, it comes down to the people, like, are they used to work remotely, are they experienced? And that experience piece is quite important because if you just came out of college, I don’t know if it’s going to make sense to join a company like Liveblocks at first, because you probably need that initial experience in an office working side by side with someone to learn the dynamics of like working at a company, and I don’t think you get as much of that at a company like Liveblocks, but if you have done this before and you have experience and you know what to do and you’re one of the good work and you’re passionate about what you do, then I think it’s a good fit.
Turner Novak:
Yeah. Yeah, that makes sense. It lines up with what I feel I’ve learned over the years. I actually have a completely unrelated question, but I just know you like basketball. So I don’t know if I’ve ever told you, I have season tickets to the Michigan University’s men’s basketball team.
Steven Fabre:
I feel like I have to come visit you then. How many seats do you have? I’m kidding.
Turner Novak:
It’s just two seats, just me and my father-in-law. We actually have... The woman who sits next to us, who has season tickets next to us, she has two seats, and I think her sister’s supposed to come. She’s like an older woman, but I don’t think I’ve ever seen her have someone at the game. It’s always empty next to us, so we talked about how we buy the ticket from her. I think she’s given us tickets before, but yeah. So, what do you think is the most underrated basketball player of all time in your opinion?
Steven Fabre:
Oh, my gosh, the most underrated player of all time.
Turner Novak:
Yeah, because you could say like LeBron’s the best, or MJ’s the best, whatever. I mean, that’s probably the answer to that question, but who do you think is most underrated basketball player, or most underrated position, or skill, or something like that?
Steven Fabre:
Oh, that’s a tough one. I didn’t think about that. Ooh, that’s a tough one. I don’t know if he’s underrated. I’m going to be very... Because I’m French, I grew up in France. I feel a player like Boris Diaw was quite underrated. He played for the Spurs, and he’s so versatile. And having played basketball, having versatile players is actually so, so helpful on your team, but also sometimes hard to fit them properly.
Turner Novak:
Is Boris Diaw. Okay. Is he retired? Oh, he’s 43 years old.
Steven Fabre:
He’s retired at this point. Yeah. He’s just drinking wine somewhere in France, chilling. Yeah. Who would be your most underrated player, then? I’m curious.
Turner Novak:
To be honest, I don’t really follow NBA that much.
Steven Fabre:
I’m big into the sports right now because Wemby is just killing it. Is just having an insane skill.
Turner Novak:
Yeah. I follow college a little bit more just because I’m Canadian, so I play hockey and fall hockey.
Steven Fabre:
Oh, didn’t know you were from Canada. Okay, nice.
Turner Novak:
You didn’t know this? I thought we talked about this a long time ago.
Steven Fabre:
Forgot about it. No, I don’t know.
Turner Novak:
I’m from Winnipeg, so I follow the Winnipeg Jets. It’s like Winnipeg’s in the middle of nowhere in Canada. And then I follow the NFL professional, and then college football, and I follow college basketball just because it’s like the local sport. And I don’t like going to football games because nothing happens. If you take the total time you’re at a football game, it’s three hours, but there’s only four minutes of anything actually happening. So I like watching football, but on TV, because being at a game, there’s nothing happening, but basketball-
Steven Fabre:
Is intense.
Turner Novak:
Yeah. And if you look away at the wrong time, you’ll miss something, right? But with football, it’s like, “Okay, cool. There’s a play that’s going to happen in a minute, we’ll talk to somebody,” and then it’s like, “Cool. Play’s about to happen.” Two seconds, it happens, and we get a minute to not watch again.
I’m trying to think of... I think to your point of, I think big guys that can pass, so someone like Jokić. Is that his name?
Steven Fabre:
Jokić, but yeah, I was thinking about him as well, but I was like, he’s definitely not in the radius because he got MVP and stuff.
Turner Novak:
Yeah. It’s why you win MVP. It’s you’re somebody who can block a shot, but can also sink a three and pass the ball. It’s just well around, versatile players. Maybe playing defense I think is sort of underrated, just generally speaking, because it’s annoying as fuck when you’re trying to score and just somebody just won’t give up. And there’s so many guys that are just so good at shooting, they’re good at dribbling, they can juke you out and pass it, but somebody who just won’t let you take a good shot.
Steven Fabre:
Beverley I think is one of those guys. People in the league really hate playing against him because he’s-
Turner Novak:
Is it Pat Beverly, you said?
Steven Fabre:
Pat Beverley.
Turner Novak:
I’ve heard of him. Yeah.
Steven Fabre:
Yeah. He’s just this smaller point guard, but he seems so... If I was that good, I would hate playing against him because he would just be running and going after you, and be like, “Damn, just give me a break.” But he has that reputation.
Turner Novak:
Because I feel most people do not try as hard when they’re playing defense. If you have the ball or you have the puck or you’re in a position to make a play, everybody’s going to try their best, but I feel most people do not give it their all when they’re playing defense. And those people are so annoying to play against because you’re just not used to it. You’re not used to somebody actually trying hard against you when they’re playing deep.
Steven Fabre:
That’s a good point. Yeah, I agree. Yeah. Well, so did you play basketball actually yourself, or?
Turner Novak:
No, I’m 5’6”. I can’t...
Steven Fabre:
Well, you could be a play guard.
Turner Novak:
Today? No, it’s too short.
Steven Fabre:
Yeah. You’d be all right. Probably not in the NBA.
Turner Novak:
I’ll tell you, are you like 6’1”, 6’2”?
Steven Fabre:
6’4”.
Turner Novak:
Oh, yeah, 6’4”. Okay. I knew you were tall, tall, but, I guess I underestimate-
Steven Fabre:
Yeah, I’m probably like a small point guard in the NBA nowadays.
Turner Novak:
Yeah.
Steven Fabre:
Anyway, I tore my ACL twice, so I’m just done with that kind of sport.
Turner Novak:
Yeah. Isn’t that how you started your first business? You got hurt, and you had to basically-
Steven Fabre:
Oh, my gosh. Yeah.
Turner Novak:
Because you were a waiter, right?
Steven Fabre:
Yeah, I was. That’s a fun story. When I moved to Sydney, I moved to Sydney in 2011, joined a basketball team, didn’t know... I had not started my career for real.
Turner Novak:
Were you playing professional or semi-professional basketball?
Steven Fabre:
Not at all. No, no, I was just playing for fun, in original, pretty good level, but not professional at all by any means. But when I moved to Sydney, I joined the basketball team. I wasn’t sure what I was going to do with my life. I was just there to travel and have fun. And so I just worked as a waiter, and making some money there, but then I played basketball game, went up in the air, somebody pushed me, landed awkwardly, dislocated my leg completely, and that sucked because it really hurt. And then I didn’t know at the time that was a terrible experience, but basically because of this, then I had to... Yeah, emergency arrived, put my leg back in place. I had to get MRIs. I had a bunch of medical bills stacking up. I wasn’t sure how to deal with this because I was overseas to deal with insurance and that kind of stuff.
I couldn’t walk, so I lost my job, and I was like, “Oh, shit, what am I going to do now?” And that’s actually what really got me into this career, which is, surprisingly, in retrospect, I’m thankful this happened in a way. But yeah, I had no money. I had medical bills lining. I was like, “I don’t know how I’m going to pay rent.” So I ended up posting an ad on Gumtree, which is kind of... I forgot what the equivalent is in US. It’s called-
Turner Novak:
Craigslist.
Steven Fabre:
Craigslist. Yeah, exactly. Yeah. Within two weeks, I had several customers, and I was making websites for them, designing websites. They were happy with it. I realized I was making more money than when I was a waiter, and I didn’t know the time that I could have a career in this, surprisingly, because I love design and I love coding, but I wasn’t sure if I was great at both.
And so, being that versatile person between the two, I learned I could do something there, and then that’s what actually got me to the startup scene in Sydney, and then from there, ended up going from one startup to the other and starting different things. And that’s what led me to eventually start Liveblocks. So, anyway, I thought that was... It really sucked at the time because, like, “Damn, what I’m going to do?” I probably had 1,000 euros or $1,000 in my bank account, and I was like, “That’s all I had.”
Turner Novak:
Yeah. Nothing like running out of money to light a fire and learn things.
Steven Fabre:
Yes. Even when you start a business, I think when you’re feeling the pressure, it actually helps you in a way.
Turner Novak:
Yeah. This is a lot of fun. Thanks for coming on the show. This is a lot of fun.
Steven Fabre:
Yeah, that was really fun. I enjoyed the conversation, and yeah, thanks for having me on then.
Stream the full episode on YouTube, Spotify, or Apple.
Find transcripts of all other episodes here.
Starting out in the world of web design sounds like navigating a tough level in https://crossyroadgame.io —you’re not sure at first if you’re going to make it, but then you realize you’re mastering both coding and design, just like overcoming obstacles to unlock new paths.