Good Enough Isn't

Why Distribution Wins: Lessons from 4 Exits and the AI Era

• Patrick Patterson

Share episode

Use Left/Right to seek, Home/End to jump to start or end. Hold shift to jump forward or backward.

0:00 | 1:12:04

Why Distribution Wins: Lessons from 4 Exits and the AI Era

with Raj Singh (VP of New Products, Mozilla)

Episode Summary

This week, Myles Biggs and Patrick Patterson sit down with Raj Singh, a serial entrepreneur who's built and exited four companies across the dot-com crash, the 2008 financial crisis, COVID-19, and now the AI wave. He's currently VP of New Products at Mozilla, where he's building an AI-powered small business suite for the ownership economy.

Together, they break down what it actually takes to build something that wins: how to find insertion points before the market does, why distribution is still the hardest problem in tech, what the "95% rule" means for AI product design, and why the team, not the technology, is almost always why startups fail.

Whether you're a founder, a product leader, or just trying to understand where AI is actually taking us, this conversation is packed with hard-won perspective from someone who's been in the arena through every major disruption of the last 25 years.

What You'll Learn in This Episode

  • Why distribution is still the hardest problem to solve, no matter how good your product is
  • The "95% rule" for evaluating which problems AI is actually suited to solve
  • How Mozilla is building a Shopify-equivalent for solopreneurs and service providers
  • Why Raj believes it's always the right time to start something and what that really means
  • The "explore, expand, extract" framework for knowing which PM archetype your product needs
  • Why teams burn out and how that's often what actually kills startups
  • What the next 12 months of AI adoption look like inside enterprises
  • How the web itself is being rebuilt for an agent-first world

Featured Guest

Raj Singh — VP of New Products at Mozilla, four-time founder with exits spanning the dot-com era through the AI wave. Builder of AI calendars, remote work tools, food communities, and now an AI small business suite (Solo, Trunk, Pencil, Post Full) inside Mozilla.

Connect With Raj

Connect With the Show

How to Support the Show

  • Subscribe so you never miss an episode.
  • Share this episode with someone navigating the AI wave.
  • Rate & review if it brings value — it helps us reach more listeners!
Raj Singh

the thing I tell people, they're still winners, right? thousands of things being released. There are. Products that are winning now, maybe the economics have changed. Maybe these products aren't meant to be billion dollar businesses, which has been this venture model, like for the venture, for the early stage venture industry. It's tough. Like how do you pick in this market?

Myles Biggs

On this podcast, we are driven by truth. Sometimes the hard truth. We believe it's imperative to be relentless for results because if you are not your competitor is we're obsessed with how to be better every day because that's what our customers deserve. And if you can set aside your ego, if you can truly be no ego, then we're the show to help you go all in because good enough isn't. Today's guest has four exits under his belt spanning the.com era, the 2008 Financial Crash and COVID Ovid 19, now the current AI wave. He's built companies in mobile food communities, AI calendars, remote work tools, and now AI products for small businesses. He is the VP of new products at Mozilla. Raj Singh, welcome to Good Enough. Isn't.

Raj Singh

Thanks, happy to be here.

Myles Biggs

So before we get into your career of what you're doing today, I'd love to start back in the early days what I could see from our research is that before college you were running the side hustle, installing ethernet cards into your friend's computers. and so that speaks to an entrepreneurial instinct mindset, and I'm just curious for you where you think that comes from.

Raj Singh

this may sound this question's been asked to me before, and this may sound, maybe even cliche, but, I went to a school that was like crazy competitive and, I was nowhere near let's say the top 10% of kids. I was probably in the top 20%. and So I, was from an academic perspective, and so I was of this like mindset of I gotta figure things out. I'm not gonna get into the top school, I'm not gonna get into the top companies competitive mindset. and, I, some people, Immigrant mentality. I'm like first generation. so my parents moved here, right? And and so I just, since basically high school, start, you, just described a good example, but throughout college was always building things. I'm like, Hey, how can I, prove it mostly to myself that I can do this right. just because I'm not like the, the top SAT score or whatever it might be, right? I suppose that maybe was where it started. And then the funny thing is if you reflect on the last five or 10 years, in many ways it's what I now most enjoy. And they sometimes say, you should take the hard path and not the easy path. But, and so now I'm starting to question this, but the easy path for me is to go build something new, which is very counterintuitive because usually the easy path's to go. take a job somewhere, but in some ways that's the hard path. I like, I look at that and I'm like, wow. The amount of preparation people do to interview. and then, just within a company, people politics and all these other sort of things that are at play, but like building something from fresh and given the network I've built because of having done that so many times, I don't care if it's accountants or legal or whatever, that's actually the easy path. And it's funny that, that sort of roll flip, which is, atypical, from what in general,

Patrick Patterson

So I have to ask what, what speed were the network cards?

Raj Singh

what speed were the network carts? Wow. wasn't that like Cat three era?

Patrick Patterson

right?

Raj Singh

I feel like 10 megabits.

Patrick Patterson

yeah. I was gonna say, I was gonna guess 10. I was gonna guess 10 megabits.

Raj Singh

So webpages are also not that heavy.

Patrick Patterson

yeah.

Raj Singh

they, they're lower resolution monitors. everything was like, but Yeah. no, it's funny because I, joke with people that the, my generation, I was born in 79, so what is that? millennial or whatnot, millennials, zoomer, or whatever you wanna call Gen Z, grew up without the internet and then had the internet. we basically knew how to build a computer and then also buy a computer off the shelf. and the same thing's actually going on now with ai. You people who knew how to code without AI and now they have to code with ai, right? But the reason I bring that up is like this sort of joke that like I always expected I'd continue to do tech support for my parents, but I did not expect that I'd have to continue to do tech support for my kids. because they're like born, like when nobody opens a computer or knows about network cards, you just buy a computer and it's fast, right?

Patrick Patterson

it's, interesting, we're in this wave and the, it's all software right now. It's all cloud, it's all that. I don't know if you saw, if you've seen, you probably have, but the, Tallas, the chat Jimmy, silicon chip that has a LLM built into the silicon.

Raj Singh

Okay. I Have not.

Patrick Patterson

so crazy. a after, after, we're done here, you should, you just go to chat Jimmy. You could do it while we're chatting. Go to chat jimmy.com. And if you're listening to this, you should, everyone should go do this. And, you just enter anything in, have it, summarize the Lord of the Rings, the entire, world, all characters or whatever. You will literally get an answer back in milliseconds. 'cause it's all, 'cause it's all silicone. So I think there's this, weird. I keep, saying everything that's old is new again, like direct mail's back and all the, but I think hardware is coming back and I think we're all gonna be buying like, AI chips and AI cards and, we're, gonna be throwing 'em in computers or they're gonna be in computers. I think we're gonna be moving to this world of hardware as the natural progression of, what is being built in software right now.

Raj Singh

I think there's been, we saw, the acquisition of Brock by Nvidia. Just LPU language processing units that are suited around inference. But what you're describing doesn't surprise me. We actually had, I think when you look at tech and look at history and you've been around long enough, you see the same patterns emerge and emerge. and so we saw this during COVID with web RTC. so video calling became a must have for everyone. And web RTC or H 2 64, Kodak, decompression, acceleration, whatnot, wasn't necessarily previously natively built into chip sets. Now it's like kind of standard and phones and everywhere. All that says contributed to just better video quality. The fact that we even have this call, but no hiccups not using a native client. And so I think, I think of course you're gonna see the same thing with inference, and with models. And, apple's obviously playing its own long game here. And we'll see how it all plays out. But, I do think you're also pointing on another interesting trend, which again, I like to think of things as pendulums. We're in a era right now, but the sort of reemergence of the thick client, because we've been going through a MacBook Air sort of era for 10 plus years, right? make it thinner, make it lighter, make it lower powered. And now we're back to sort of 48 gigs of ram and loaded, CPU and whatnot because we're actually running things locally, which is, there was actually a, interview I saw, I forget with who, it was the, one of the PMs from Gemini, talking about how, they had built, they had made a range of bets, obviously with, Gemini and around their coding tools. But one of the bets that they felt they miss, segmented from a positioning perspective was they were still betting on a web-based experience cloud sort of thing, to drive it. And Claude kind of bet kind of local CLI offline. And it turned out the latter is the current winner. Now, of course, the, tech cycles are so fast and compressed, this could completely shift within six months. At least Right. now. these little positioning segmentation exercises that you do when you build new product are really important early on because, they can change very quickly, but they can also make and break, who are the initial victors?

Patrick Patterson

Yeah. Opus 4.5 came out in Thanksgiving ish, and Gemini 3.0 and what was it? Codex 5.2 all in the same week or whatever. and I immediately jumped in and started creating web apps. It's the first thing I did. and, too many web apps and pushing 'em up and, up to Sal and, getting 'em hosted and doing all this stuff. I'd say the past, like month and a half to two months, I have, converted almost 99% to local apps, locally running apps, CCPs. and I now have kind of the suite of apps that I've created on my own computer, that is more, more controlled. the model really, if it is AI powered, the model becomes something that, that I can switch out either running something locally like Quinn, 3.5, 27 B, or, switching over something like Gemini or, Claude. But the, it's this ability to, I think, I think that's right. you have people out there buying $20,000 MacBook Studios now running these local models and, running these, things on it. And then again, that will then. everything will swing and then that will be in the cloud and then we'll go back to the cloud 'cause it will be cheaper and then it'll back and forth. Right.

Raj Singh

I think one thing that's unique about this, one of the motivating factors to run a thick client is the web itself has not, it pains me that the web, the crypto community tried to take the term Web3, but if we call this new era Web3, which is trying to redefine what the web experience might look like in an ingen first kind of universe. The web has not been, rebuilt in that way yet. And what I mean by that is for those old enough to browse the mobile web in 2008, 2009, when you used to open a mobile website, it would say View Desktop site at the bottom of the mobile website, right? So there was like two versions of the web, right? And you're starting to see that now with agents. You could have an onboarding flow that's oriented around agents. Look at Malt Book, the sort of social network for agents. People are dropping in links there. If they click, it's an agent clicking that link. Now let's give them an onboarding flow, which might look a lot different to a, human optimized onboarding flow. and just like you're seeing support sites now have a human readable version and then maybe a markdown format, for an agent, Right. So I feel, one of the things, that's happened is the emergence of, agents controlling your computer or using a headless browser or whatnot to execute tasks on your behalf. So take Claw, for example, or quad cowork. The web publishers can't really block that traffic 'cause it looks like client side traffic. But if you run that in the cloud then you know, it looks like bot traffic and so it's much easier to block. And so I think this is one of the other sort of pressures that's driving, a thick client approach and running it local because it looks like just regular human usage, but it's actually your bot and agent. And of course this is gonna wreak havocs with or havoc with what's gonna happen with ad revenue and reporting and general

Patrick Patterson

Yeah.

Raj Singh

analytics and those kinds of trends. But I think that's one of the reasons that's pushing the sort of, big client local approach, until we get to a point where the web has evolved in such a way to allow agents to control and use it on a human owner's behalf in the cloud. But until that happens, at risk of getting your IP blocked, you pretty much have to run local.

Patrick Patterson

Yeah. I, think it's interesting and you have, not someone that I could talk to probably has the firsthand knowledge of where. Genic browsing and versus human browsing is going then, the guy who's running 70 products at Mozilla. Firefox has always been that, that that solution out there that, thinks differently and builds something differently and builds, be honest, maybe building the, browser. I, I wanted when I was using different browsers. That's always how I've, described Firefox, and, how are you thinking about now going forward? And, I know Myles you wanted to talk about the past, I wanna talk about the, for the future, but how do you think about going forward? Like how is the browser gonna change? 'cause we've had. We've had some fits and starts of people trying to change the browser, right? And trying to make it an genic, experience and more AI forward and, all of those things. How do you see that changing going forward as, we go forward? Is it a, like desktop and mobile? Is it, all of these things come together and it's a seamless experience? Is it separate experiences? What, are you seeing?

Raj Singh

I think there's a lot of different hypotheses depending on who you ask, and some of those hypotheses are not. Entirely neutral opinions. They're obviously have some form of bias based on where they want their company to go or be. one thing we should separate is the, there's like the role the browser's gonna play in this new sort of AI first world or agent first world or whatnot, but forgetting that the users don't necessarily move at that same pace. And, we've seen over the last year, multiple ai, first bespoke browser experiences that have emerged. But from a market share perspective, they're all quite small, right? They might be a 0.1%, 0.2% at best. they're not moving anyone's needle. And it really goes to show that we have to separate distribution from, the sort of ai. First you could be the most innovative browser, but that doesn't mean you're gonna acquire any market share, right? I think, as I alluded to earlier, I think of the web as the new API layer of the internet. There was a time you needed APIs, but now if I can have an agent execute tasks on my behalf acting as a human now, it doesn't matter if there's an API there or not, right? I can still get stuff done. And so I think this is just a complete transformation on how we do work. And I, one of the things I've been pushing even my, within my own teams is this thinking around systems thinking. Because this is just a different way of thinking, think about systems and think about reusable pieces that we can build and reusable workflows because everything has the opportunity to be automated in some way or form. right now, I think a lot of what we're seeing, and again, I'm speaking holistically, I'm not speaking Mozilla specific is a lot of experimentation. so obvious things, obviously, let's put a, let's put a chat bar on the side of a browser. Sure. these are table stakes sort of things, but what's next? Am I gonna execute tasks from there? are there common workflows? And there is, an 80 20 rule, right? Just like Siri, it turned out playing music and asking the weather turned out to be like 80% of the calls. And so, they moved some of that stuff into a local model, at least the voice to text part, right? So just to speed up the response time, the initial response. so I think, I think, I don't think people know exactly what it's gonna look like. for sure the sort of Chromebook argument, oh, is the, is the browser gonna run the os? People have been saying that for years, but in many ways I feel like Chromebook has not had the outsized success that you would've expected. It's, certainly had an impact, particularly in education. But now we're seeing it swing the other way, right? Like really, thick clients, and full blown os and very hard for me to predict, playing that role. I think as Mozilla, our stance has been a lot about user choice. That's always been sort of Mozilla's position. and there's a lot of people that don't want AI in their browser. They really just want a web where they can browse as opposed to it trying to be proactive or personalized or summarized or things like that. So we're seeing all of the above, and I think these things take a lot of time, chat GPS out there, but Google's market share hasn't actually gone down. I think maybe it's even gone up like in terms of overall, search market share, right? Maybe it's like maybe within, within 1%. and it just goes to show that, these changes take a long time and distribution is a lot harder than people think.

Patrick Patterson

Yeah. I love this idea of the browser, the web being the API layer of, getting work done. 'cause was actually, a great example. There was a website that was super hard to use built on top of Salesforce and, it was built with, Node or whatever and, and all the API layers were there underneath powering the front end. so I had a, I had an agent go in, figure out all the API calls, I built it into an MCP, and now I'm just in cloud code and I can get all the information I need from this website without ever going to the website. If I need to go there, I can still go there. But it's this, idea of now. was, and that would've been impossible. not impossible. It would've taken a dev team way too long to make that project even worth it a year ago. And we're now in a world where I can do that in an afternoon and then have something that's infinitely more reusable than I had before. And I think this idea of more we can get standards like that, and I don't know if MC P's quite the solution, and I know we're all moving to CLIs now. like the, more that we can get standardization around how can either a human use it or a cl a an agent use it or an automation use it, to get work done. how can that be the enablement layer? I think that's really fascinating.

Raj Singh

I think this notion of Claude as your canvas or really the terminal window as your canvas or chat as your canvas to everything. I feel like when ChatGPT first came out, there was some questions around, is this the modality to rule it all? I don't, I'm not just people who are in a canvas that everything's gonna be driven through chat. I'm not there yet. I think there's plenty of examples where. A gooey, a wizzywig or whatnot. It's just a much better experience. it depends on what you need to do, whether you need to collaborate, it depends, right? But there's a lot of workflows where you can, you just can't achieve the level of personalization or you just can't communicate the intent or what you really want without writing it down. whether you're using voice or text, that's independent. we're seeing, at least within my own teams, a lot of bespoke apps being built. people using, cloud code to build interfaces into existing, Mozilla applications, maybe that are currently available through SSO or whatnot. they might be plugging into an MCP, they might be going through an API. It's all in the above and it does raise new questions. 'cause you have all these bespoke apps. I'll give you an example. And again, this is not a real example, but just a hypothetical. Using, cloud code to control expensify or control your expense management system. Now you have one interface, somebody else has another interface. there's a certain, there's a certain dopamine hit from like building it and owning it. And so in some ways, agents buil, but people don't necessarily want to work on other people's agents. Everyone wants to build their own. so everything is bespoke and, it's not clear. I think there's multiple things that have emerge, right? Like this whole, whole sort of shadow IT concept. Like where's all this data flowing? what are you plugging into? Is it going into places it shouldn't be? then who's managing all this? So I think this was called out in a recent, Twitter post I saw about, the only four jobs that'll be left in the second one was all related to Cloud infra, SRE. Yeah. of course. I mean like the, amount of applications being made available on the public web, through Netlify or cell or whatever you're using, it's wild. And, are they secured and are we sure they're not revealing any internal data and things like that, right? Because. Cloud's not doing the info right. I think, there's a little bit of wild, west going on the types of things people are doing, people using open cloud, with WhatsApp, with work stuff, and with their email. It's I'm, seeing all kinds of crazy use cases, but in many ways it's also exciting because I think this is the time the best practices are gonna emerge. and of course the pace of innovation is all going so fast that these problems will be solved. We're I call this the, we're in the, build your own computer era, of, agents, right? Just just like with computers back in 98, but like within five years, six years of, I think literally by 2003, 2004, nobody was really building their computers anymore. At that point, everything was just buying off the shelf. So we're, gonna be there in a much faster cycle 'cause everything's compressed. but, a hundred percent agree with you. there's, A lot of bespoke apps being built and people connecting things and not recognizing the compute costs. DOS attacking themselves 'cause they're running crazy database queries through their MCP, return on token spend might be off the charts. there's a whole bunch of, there's a lot of stuff going on right now and I think people, enterprises and people are just trying to figure it out.

Patrick Patterson

Yeah. and I wanna dig into how you take all that and lead a product team, in 2026. And I, I think that is super interesting and I'm super fascinated by that. But just to, it's very fascinating to me. And we, we've rolled out, Code, I think eight, four 40% of our staff now, here. And we have a lot of folks that are, building apps. but they, there was someone who, they build an app and it goes through all of their emails, all of their Slack messages, all of their Asana tasks, pulls it all into context. writes a, a, six point. Bullet summary and then sends them a slack message in the morning. And like you just spent 375,000 tokens to get a five bullet summary of, the tasks that you need to do. And, where maybe a WY wig, maybe a, front end, maybe an automation that is more done programmatically would've actually been better there. and so there's this, the, there's this weird thing happening where I, do think, and I don't think it's gonna be this year, I think it's probably gonna be closer to the end of next year. We're gonna start having, and I'm having these conversations and people like you are having these conversations, but token optimization is going to be just a huge part of what we're doing as these bespoke apps come in as everyone gets access to these, tokens have gotten 1000 per, 1000 x cheaper over the last three years. But it's still I still think we're gonna have people that are like. Hey, why are you spending half a million tokens to send you a text message? and does that make sense and is that the right model to use? And then, as people get better with it, utilizing better prompts and, better context to use cheaper models to get the same work done, which I've had a lot of success with. that's gonna be a, major part of especially large enterprise, that is adopting this on a large scale.

Raj Singh

I think for most enterprise they're prob, what I've heard is numbers in similar ranges in the forties in terms of a adoption of cloud code within the enterprise. what's interesting about that number is, the average IT organization, a mature IT organization might be 30 ish percent engineering, right? So that means they've achieved broad based adoption of engineering. I'm just assuming. But they've also managed to get early adopters across other functional roles, which is great. But if you're an enterprise lead, if, you're a CEO of a company, the first thing you're thinking about is how do I get this from 40 to, let's say 60% or 80%, right? And so the last thing you want to do is start putting guardrails all over the place. I'm gonna limit your spend or limit your experimentation or whatever. I just want you to start playing with it because a lot of this is ex learn by doing and trial, by doing and experimenting. And yes, it's hard to keep up. It's moving all so fast. And your sort of unique approach to something would be completely different in three months and who knows? But I just need you to start somewhere. And so I feel I, on one hand, I agree with you, at some point the ROI calculation is gonna emerge, particularly with smaller companies. Maybe they're already close to 90% adoption across their different roles, right? But in larger companies, we're gonna be in spend mode, for a while, and just trying to drive up that adoption. Obviously there's some responsible AI use and things like that around that, but. I do think at a certain point, those questions will emerge. And I also think the models themselves or whether it be the models or whether it be orchestration layers that are built around the models, the scaffolding or whatnot will get better at knowing Hey, this is the type of query we should run through an MCP. We should run this just through, through the API. This is a type of query. We should use this super distilled model because it's like very cheap. and this is the type of query that we should send into some sort of deep think and use some like auto research, DPT or whatever it's right. And so I think I think it depends. but this stuff is just being figured out now. But you can imagine that most of this will be solved in a matter of a couple years.

Patrick Patterson

Yeah. Geminis, they're not doing a great job of it, but Gemini CLI does a, pretty good job of switching between four or five different models with 1 1 1 1 query and using 2.5 flash for tool use and using 3.1 preview Pro for the deep think and switching back and forth. say outta the, out of the three major players, it's probably the best at that model switching, to reduce cost, also increase speed, maybe at a hit to output. that's the trade off that you're always gonna make, right? So you take all of that, you take this world that we're living this uncertainty of over the next year and a half, how do you lead 70 products across Mozilla? How do you even know what to seven 70? yeah. how do you lead seven products across Mozilla? and know what to build, when to build it, when to get people excited and experimenting, when to get people. it's, it, to your point, it is super exciting to do the first 80% of the project that takes 20% of the time and get that dopamine hit and launch and ship a, ship, a project to yourself. it is not super fun to then take that and, distribute that or get it into a place where it's easily distributed and it takes a lot of time and effort to know right from wrong. So how, are you navigating that world,

Raj Singh

So I, yeah. The funny thing is there's a lot of hypotheses around what AI is good at, and obviously AI is a thought partner, amazing. But I do think some things haven't changed, and I don't think AI is telling you what to build. AI may tell you how to build it. It may inform. Own thinking on what to build. But I, like to take the opinion that all the information you consume, whether it's customer feedback, survey data, quant, qual, I don't care, ai, those are all data points inform your own model. And then you have to think about it and say okay, what's the next move? if everybody followed ai, exactly right. If AI became let's say we were artificial super intelligence, a SI and we were just executors of what AI said, right? everything, there would be no winners, right? Everybody would, there's no edge, there's no, So at some level, there is some human judgment, human creativity, human taste, whatever you wanna call at night. I struggle sometimes with these terms 'cause I don't know how best to define it. But, there is original thinking. and, as a sidebar, that's a whole different topic, just my concerns around that, right? I've got, I've got a son who's a teenager, right? And like, how much thinking does he do on his own now versus lean on ai, right? And how that affects the brain and development whatnot. But in any case.

Patrick Patterson

I call it taste, by the way. I call it taste

Raj Singh

Yeah, I, mentioned the word taste, right? Yeah. yeah. There's a lot of, there's a lot of words for this. I, when it comes to building product, I don't think a lot has changed. I think, I think the big thing, which has remained the challenge since let's say 2018, has been distribution and growth. I think if you look at growth on a 25 year arc, you will see these interesting patterns. cold email was very effective in 2002, 2003, right? tapping, the Facebook wall, very effective social in 2007, 2008, right? Mobile app store is getting featured very effective in 2010, 2011, right? So if you played this out, ai, SDR, what we call now, but in 2018, very effective, like LinkedIn dm, cold outreach, personalized. Very effective, right? Very high conversion rate. All, every channel is flooded now, right? Because through automation, ai, go to market, et cetera. So distribution has never been harder and development has gotten easier, right? So the volume of apps is just skyrocketed. And so I think, I think one of the things I've been telling my teams is I think there's too much focus now on identifying the GTM stack, which is everything from like SEO to social media outreach, to programmatic blog, programmatic social, to, paid acquisition, whatever. And we need to lean more on trust. and basically PLG, like user referrals and word of W, word of mouth optimization and things like that, right? So I think everyone is oh, I can run all these things through, open claw or whatever, right? So I'm like, let me go automate all the GTM workflows, but everyone's automating all the GTM workflows. So I don't know if there's any edge, right? Who knows what percentage of Reddit comments are actually human authored at this point. even with moderators there, right? I think, that is something I'm definitely still thinking about, right? Like evaluating an idea, can we get reasonable distribution? And, there's weird moats too, right? Like maybe the old school's, new school again, right? So we're thinking like consumer all of a sudden is hot. Why? 'cause consumer always plays on a vice and consumer tends to be more PLG driven, versus like traditional SaaS. consumer has natural moats because of brand retention and whatnot, right? Whereas so much of SaaS work has been personalizing a workflow, which plug is now very easy, you can replicate, right? maybe top down big enterprise sales becomes hot again, right? Which is. Been sort of an antibe category for so many years. 'cause he wants to do like outbound in person oil meetings, to sell the big enterprise. So, I'm, just saying, distribution is something I'm still like always top of mind. Then there's a whole bunch of other things, in evaluating an idea, outside of distribution, I, I break it down. there's domain expertise. do we know about this space? Do we have an insertion point? Do we have some sort of, what I call creative execution? do we have an angle of attack? Is it ignored in some way? But in, in some ways, with my teams, although we're build looking at things from an AI first sort of perspective, we're not building AI developer tools necessarily, or we're not, leading with ai, right? Like it's, AI is a core part of the experience, but it's not, it's like Google photos. Like you don't even know it's using ai, right? But it was doing amazing AI well before the sort of gen AI phrase. And part of that is, there is a little bit of fear, I. if I'm entering a category that Claude could just build in natively in the next six months, that's a problem, right? So that's become a new question too.

Patrick Patterson

Yeah.

Raj Singh

can I just do this with Claude? And what's my advantage? What's my disadvantage? some people are describing this next era of software as services driven. So we're real humans are involved in the work, right? for example, it may not be about social media automation, it may be about white glove social media management, right? so it's versus social media, DIY tools, right? it's, I think there's a lot of factors to think about. In the end, there's a little bit of intuition. It's the people. it's knowing, when to shut things down or keep pushing forward. But I would say if everyone was so good at this, like we wouldn't have venture firms that are returning money and a lot of venture firms that are not, right? It's just hard, right? So it's not easy. There's new challenges, obviously, even within a bigger organization. But I think, pattern matching is a large part of it. I've been building products my whole career. I don't try to focus on areas I'm not good at. I've been very consumer focused. and I feel like I have some edge there in thinking about consumer psychology. 'cause I think consumer's all about human psychology. but, I don't have a magic formula like this is how we green light something or not. But I certainly am thinking about a range of things, and that's probably a whole different topic in itself. Like all the sort of.

Patrick Patterson

No. Yeah, I think it's, I think it's, I think that's super fascinating, right? And this, idea of being more services driven, we're a services company, right? and I look at some of the other out there, like specifically like travel agencies. And you look at what happened to travel agencies when just like Google Flights came out or whatever, kayak or whatever it is, and they were decimated, right? They were decimated. About 95% of them shut down, but they didn't go away. The ones that survived were the white glove, were the high touch, we're the really, we're gonna go above and beyond as a travel agency. 'cause there's still a need and there a time arbitrage that happens no matter who you are. and I think that will still exist when we're all given the same tools. Then how do you, how, what is your edge? How do you differentiate? And I think always been the difference between a great company and a company that goes outta business. Is that edge is how you stand on top of that tools, not how do you utilize the tools to send more, spam messages via LinkedIn, right? think like it's, going to be that human in the loop, the, human taste. Now I do think what the difference is, we're all standing on a much higher platform. Than we were. And so the difference between good and great is gonna be hard to spot, right? and I think that is where we're all gonna need some education and a across the board on, oh, is this, what is the value of 5% better because of humans there? and I think that understanding what that value be interesting.

Raj Singh

And I think the thing I tell people, they're still winners, right? thousands of things being released. There are. Products that are winning now, maybe the economics have changed. Maybe these products aren't meant to be billion dollar businesses, which has been this venture model, like for the venture, for the early stage venture industry. It's tough. Like how do you pick in this market, right? Late stage, at least you can bankroll like, okay, hit a certain in, certain revenue threshold, something will materialize. I can structure in a certain way, whatever, right? But like early stage, it's tough. but, here, ignoring that for a second, there are still winners. There's gonna be winners, there's gonna be losers, and, there are plenty of acquisitions happening. So there's plenty of liquidity in the market, right? I just think this is, really interesting times. I've always been in the camp, people like, oh, it's a great time to build. And I'm like, I've been saying that every year, right? I had a startup through the.com crash through GFC, through COVID, th three of the big financial crises of the last, two decades. it's always a good time to start something and this is no better than any other time. And my general opinion is yes, this feels like a, very disruptive moment, akin to the iPhone being released and how disruptive that was like in terms of mobile transformation. But it's gonna settle. this will become the new baseline. and there will be opportunities and edges and insertion points and areas for disruption. So I think it's, I think it's just, as a new products team, I think it's just an exciting place to be. 'cause it's like the world's your canvas, but it's also, it's also tough, right? You got all these different options.

Myles Biggs

I'm sure not every, new product you put out is a winner. You mentioned there's winners, right? So is there one that recently it was like, it was a great idea, but it didn't work. Maybe it came down to distribution or something else.

Raj Singh

Yeah. it's hard. One of the hardest things I feel like in building a startup. Is knowing when to pivot. And there's different terms like pivot versus hop, right? So sometimes pivot is just like same product, but changing the go to market or changing the segmentation a little bit. Hop is like throwing it away and moving into something entirely different. Like the transition of from games to Slack, right? Like with that backstory, that's definitely hop, that's beyond a pivot. knowing when to pivot, it's hard, right? It's a typical startup, historical, liquidity, seven plus years. and I've built enough products in my career that I think it takes about three years to even see if something's gonna work, which is, that's a long time, right? and it's a tough, it's intuition driven, but you wanna make sure you give products enough space and enough capital so they can run the experiments that they want. Oftentimes things don't work, not because of the product, 'cause the team gets tired, and. They mentally are exasperated. But, I don't think there's a magic bullet answer here. And I once had a board member that said, I've shared this quote before, but I, I'll share it again. It's, we never make the wrong investment. We just make it at the wrong time. It's a hell of a quote, because it's, it's wow. but at the, same time, it's really calling out the reality that, sometimes that is just timing.

Myles Biggs

So when you say the team gets tired, I think that's really interesting. We focus a lot on the technology in this conversation, but that are humans, like we've alluded to, tires the team the most, would you say?

Raj Singh

building a startup is, lots of ups and downs and I think, I can speak at it from a Mozilla perspective. you think about a bigger organization, different types of companies attract different people. So if you think of mature organizations, you attract certain kinds of people, the fear of being wrong is like a big deal. it is how these companies reward and in startups you're, you have to be very comfortable being wrong and it's easy to say? that, but it's hard to live it. and I think, I think going through those emotions and going through the sort of work of there is no clear answer 'cause you're always dealing with incomplete data, incomplete time in complete resources. I think it's very hard. there's a framework I use, which is called the three X framework. It's from, I forget his first name, but Beck, he has this, he talks about different, like PM archetypes or leader archetypes. And he says in any sort of product cycle you have the. Explore phase, expand phase and extract phase. So it's three x the archetype of a PM in the explore phase. What they call zero to one is very different to somebody who's in the expand phase or extract phase, which is much more like an operator. In that phase, you have tons of data and you're really trying to run hundreds of ab tests, ex out a 0.1% improvement in the drop off a particular screen or drive another 0.2% improvement in some conversion funnel or whatever it might be, right? When you're in the explore phase, you're, you make a change and you get a 50% improvement, right? You're dealing with step functions, you have very little data, and you can imagine in an organization, everyone's a backseat, driver, right? everyone has a strong opinion. Everyone's, when things aren't working, everyone wants to get their feet, hands involved, right? So the whole, all of that can be very exasperating, right? And so in some ways that perseverance, the sort of, founder mode mentality, whatever you wanna call it. it's hard. It's it's, and for a lot of people, just dealing with the emotional ups and downs and constant good news, bad news, and always running in a high stress mode with a lot of anxiety, burns people out. And we see, and, that's why I actually tell people the, the number one reason startup fails, they run outta money. But the number two reason is 'cause the team gets tough.

Patrick Patterson

Yeah, I think it's interesting, right? And just real quick on that, there's this, we've been around for only 16 years, so it hasn't been that long. and it took us a long time to become an overnight success. And we're still a very small company, right? and, one of the things I, it was a, few years back, sitting with one of the members, a member of my leadership team talking about, yeah, we have this big problem and once we solve that, we're, it's gonna be good. It, we looked at each other. It's no, they'll just be the next big problem and they'll be the next big problem. And when you're running that, when you're running a business, when you're, when you're fighting for it, when you're growing. you solve two problems. You create three more. They're bigger, they're harder. and that grind is a lot. Like you have to want to do that, right? you have to want to fight up that hill every single day. I can see, you've been running a marathon 26 months. and then it's okay, great, you just finished the marathon. Now let's go run a a, triathlon right after it. and then let's, and then you finish that and it's okay, now let's let's go up Kilimanjaro.

Raj Singh

It's not for everyone. it's, there's roles for everyone. but doing a start or founding a startup is not for everyone. And of course, you raise venture money and then the treadmill really starts now there's pressure to run faster, go faster, grow faster, fail faster. It's in their best interest for you to fail faster than it is to linger. and so they'd rather just move on, it off, right? And refocus their energy. And but from your perspective, this is your entire brand reputation. This is what I mean, what, like that sort of fear failure, like when you're in a big company, it's very easy to say so and so didn't work because of X, Y, and Z. Or, I, have to make this decision because leadership is saying X, Y, and Z. But when you're in a startup, like you can't blame anyone else. Your brand is your startup. And it's hard. A lot of people struggle with a lot of that. And It definitely, I, do subscribe to the thesis that being a founder or CEO for enough years, you do become a little bit sociopathic, because you lose your empathy. you're empathetic to users and whatnot, but there's just almost nothing you haven't seen, like you've experienced every, like it is very hard to get me super excited and it's very hard to get me super upset 'cause it's like I've seen, it's unlikely I've, I'm experiencing something worse than a scenario I've been through before.

Patrick Patterson

That is so true. Like I, we like losing a large account or something, used to be devastating and now it's oh, okay. let's give ourselves five, five minutes to be inbound about that. Move on, do a retro, figure out why we did it, fix it going forward, and let's just, move forward. And I, give that response to people that come inside the organization and they talk to me and they're like, are you not mad about it? I'm like, at what? At what point? like there was a point in 2017 where we almost went out of business,

Raj Singh

Yeah, exactly.

Patrick Patterson

What are we talking about?

Raj Singh

I had Crazier moments,

Patrick Patterson

Yeah.

Raj Singh

and it's very difficult. I've had moments where like incredibly rough news for the company five minutes before going into a standup where I have to cheer everyone up, right? So like you have to operate in a mindset where you can really like box different pieces of information. And I don't think everyone can do this like it's a. It certainly battle hardens you.

Myles Biggs

I am curious

Patrick Patterson

Yeah.

Myles Biggs

how you do it. 'cause you've done it for many years, and I see that you do, Brazilian juujitsu, so maybe that's part of it, just having that physical outlet. good for the mind, but how do you not get tired out?

Raj Singh

Of course, I get tired. so I've, between startups, will often take an extended period off, in both whether a good outcome or a bad outcome just 'cause I need to rest, and decompress and disconnect. there's no doubt about it. I, for a while I used to, like I re a lot of, I think for me is reframing thinking. So I used to think of stress as a bad thing. Then I this read this book about good stress and I was like, and I reframed, I'm like, oh, stress is a good thing for me, right? so I started reframing how I think about things and that's helped a little bit, but I also don't, I would say. I still work hard. Like when I think about my startups, like I still, I'm still pretty much always online, and we're working long hours, but like the way I used to work where it was all the time, but without thinking about what's important and now I have a much better sense of what's important, what's not, I feel like I'm able to juggle and handle that a little bit better. I think the other thing too is I've gotten better at knowing, 'cause one of the, one of the biggest unlocks when running a company, I feel is the moment where you can take a vacation and actually put your phone away, which is really hard. It takes a long time. I feel like when you hit that point, that means you not only hired a team, but you've hired a team where you're willing to delegate decision making too. Because this is a level of trust and judgment, right? And so I feel, that's a big milestone and often doesn't happen till like year 3, 4, 5, right? but when you can get to that point where you can feel like you can disconnect for a few days. That's really nice. So I don't, have a slam dunk answer, and I would imagine this is different for each person. but, without a doubt, I definitely go through moments where I'm like, I need, disconnect. I'm exhausted. There's too much going on. And now it's, honestly, in a weird way, the pace of, ai, it's recreating this feeling for a lot of people where they're like, like it's constant anxiety. are my teams moving fast enough? Are we doing as much as we can? Do we need to invest more, do more? And so we don't know. and that creates this anxiety and there's just only so much time in the day because we still have people management, we still have all these other things that are going on, right? We still have to, we have to still have to steer the ship together, right? So it's, it's definitely, a challenge.

Patrick Patterson

Yeah, I think this the, I think it's real. I wrote an article a few weeks ago on LinkedIn, but, this, about my addiction to AI and, just giving it a name and, going through that, but this fear of not innovating, I think is real. it's, it, got to the point for me and, just talking about it and writing the article was actually, part of me figuring out what it was. But, it got to me where I was in a meeting and I'm like, wasting time. we've spent 10 minutes arguing about what this slide could look like. I could have built, I could have built 10 things by now, right? and it's the latency of human interaction when you have instant. and I'm a DHD, so I get, have 29, terminal windows open across three LLMs, and I'm, I'm building all this stuff at, the speed of thought. and then to, then be slow down do the things that are important, be with people and be with your family, and connect with your friends and disconnect. those are important things for who you are as a human to be. Then challenging that through this fear of not innovating, and then compounding that with this founder mode, grind mode. Gary V you gotta be working 18 hours, 20 hours. wanna outwork you, you combine all that together and, and the, best way I could describe it. I was talking to a buddy after, I wrote that article, I was like, remember that time when video games were a thing and you got addicted to video games and you would play video games, and then someone in your life would be like, you're wasting your life away because all you're doing is staring at a video to game screen and playing siv or whatever it is, right? every, if, you've ever gotten into video games, you've had that moment in your life, and this is that moment, but everyone's cheering you on because you're building something, right? And so it's a, very interesting, a very interesting world to be in, and I think it's real. And I think that. what you described and what I'm describing, like that's gonna be a thing that people are going to have to start to deal with in three to six, to nine to 12 months as it gets rolled out more. And as we figure out what, how people should work and what is acceptable amount of work and, we start to normalize again. But this is like a real psychological moment for a lot of folks that, that I think needs to be, we need to figure out how to get through it,

Raj Singh

A hundred percent agree. I dunno. I think you, it's fine. I x-ray or zoned in on that comment. You had 29 terminal windows open. And then I'm just thinking through and the ROI of all of these sort of things and being, one of the things that I think is really hard in building new product is zooming out. because you get so stuck in the weeds, and I feel like the dopamine hits of running all these clot instances, doing all these things for you, and like running your whole org chart through Claude, you forget to zoom out and think about strategy. And I actually think that's an area where AI is pretty bad at, is, thinking about unique strategy to win. and, so I forced that in my teams. Hey, take a moment. Take a step back. or doing the right things. Is this still the same way you would approach it, Particularly early on when you think about zero to one, you revisit your strategy every few months, right? Like when you become the more mature you are, you might revisit every two years, but when you're small, you can pivot very fast, like turning a motorboat, right? yeah. But, I agree with everything you're saying. This is a, such a transformation transformational era. and it's created a lot of new pressure on our teams as well. Like maybe they're not writing code, but they're reviewing code at levels that they've not had to review before. Our SRE teams deploying infrastructure at levels they've not had to deal with before, designers and whatnot. So across the board we're seeing like, just a lot of pressure, a lot of volume. Yeah.

Myles Biggs

Yeah. Yeah, that's something, I've had my own version of it. I am, to the two of you talk has been fascinating. 'cause you're throwing out all sorts of acronyms that, I'm not a developer I've been dabbling. I, spent time with Patrick in person like two months ago and he blew my mind, me all this stuff. And then I felt so far behind that pressure and I came back and started billing stuff and I get the dopamine hit for sure. And even this weekend, I was on my run and I had an idea, and literally while I'm running, opening up claw on my phone and like talking to it about a skill I could build to do something and then finished it when I got done. it's this feeling of you're, you can be always on in a way that you couldn't be before where it feels productive. But like you were saying, Patrick, it's like at what cost? Like you're not making those human connections and you feel that pressure. So Roger, I'm

Raj Singh

and does it actually move the outcome? That's the

Myles Biggs

right.

Raj Singh

thing. I, I tell my teams, I'm like, we're doing all these things. But like in the end, is revenue going up faster? that's, really the, at the end of the day, that's what's the ultimate grade of the business? there's a lot of things obviously, but revenue is certainly one of the top three or four things.

Myles Biggs

Yeah. Which speaks to the way, kinda what you were saying before about reframing, right? And do you find that's a way that you deal with pressure, help your teams deal with pressure? It's just those quick, someone tells you a story like I did and you, have to flip it on its head for them to get them to think about it a different way.

Raj Singh

Yeah. I, think a lot of problems can be solved or a lot of stress can be solved by just reframing your thinking. So you take that, question and then is there, can we just frame it in a different lens or from a different angle or a different perspective or a different archetype or whatever. And, and it's very out of box, but I do think it has a very de-stressing ability because it allows you to, just think more clearly. The volume of all the dopamine hits from all the cloud terminals that are running and all the things being created is in some ways stimulating more a DD mindset. And so I think it's hard to think clearly, and take a step back and stop and just think, we're forgetting how to write from scratch, right? If you think of the Amazon PR FAP model in terms of building product. these things are still critical and important. And so I am curious to see the implications of these sorts of changes on the next generation of builders. in the meanwhile, there's gonna be lots of, green opportunities to make money. are businesses gonna be billion dollar businesses? Are they gonna be small? I don't know, but, definitely, the pace of change and transformation feels like it's at record speed. While at the same time I'm still, trying to, part of me is still trying to go through a little bit of when I made that transition from coding full time to, let's say, doing more product stuff full time, I felt I was gonna lose my IC skills. Now I'm going through it again and I'm like, am I losing my own ability to critique design? And just leading on AI devalued everything. And, but, and and now, I have to justify why I want a particular decision because the AI said otherwise, not because of another human, which is really interesting. right? So like these things, ai, another

Patrick Patterson

Yeah.

Raj Singh

stakeholder at the table.

Patrick Patterson

That's really interesting. I'm sure your answer to this is they're all beautiful and unique snowflakes, they're, amazing and you love all of them, but, the products that you're putting out into the world right now and building. Which are you most excited for and why? And you know what, I go to Mozilla's product page and say, is the, thing that I should download, and everyone knows about Firefox, but maybe it's Firefox, but what are you the most excited about over the next 3, 6, 12 months from your team?

Raj Singh

So we're, there's lots of exciting things happening. everywhere. and even within lots of interesting things we're building. I have a suite of products that I'm building for small business owners. and I am, I think, I think what we're seeing, I call it the ownership economy, this sort of, solopreneur, whatever you wanna call it, this is an interesting time. People wanna be their own boss. AI for the first time has enabled them to do that at a level beyond anything we'd seen before. And the last time that happened was COVID with remote work, people starting to spin up their own projects and passion, right? I think there's a lot of opportunity to build disruptive products for all sorts of workflows, whether it's. Them getting started with their website. We have a product there called Solo, whether it's, back office accounting. We have a product there called Trunk, for the soer, whether it's your virtual receptionist, we have a product there called Penciled. whether it's managing your social media, we have a product there called Post Full, right? So we have ranges of products for different, I'm just going through a handful, but for different pieces of the workflow of a small business owner. and, effectively trying to automate all of it into a sort of single suite. so you can just go off and going, Shopify did that for e-commerce, but I don't think there's a company that is going single handedly like the service providers, the plumbers, anyone who builds by the hour, right? The electricians, the coaches, the tutors, whatever. So I think that's the opportunity and that's the space we're playing in. It's very aligned with, Mozilla and enabling entrepreneurs and enabling solopreneurs to build more, right? that's an area I think just in general I'm very excited for. And part of that too is just, I think. Those trade occupations. if you looked at IES Repath, leaked research around, what jobs are gonna survive in the AI era, those aren't getting replaced anytime soon, right? So in some ways those are some of the fastest growing segments, in terms of occupations and they need a lot of tools.

Patrick Patterson

Yeah, about 30 per 30% of our clients are in the EDU space. And of that, I'd say 60 70% of them are in trade schools. and it's interesting. I've been in the EDU space since 2005, so 21, 21 years now. and I remember this, there was this whole for-profit thing that was happening and, trade schools were looked as less than. And, and now there's the, I think, my, my, nephew asked me what he should do, and I was like, you should be an electrician. more than anything, go be an electrician. and he is like, why do you say that? I'm like, I try to get one to my house and it's impossible. So there's gotta be, that's, my, the whole data set that I have. but, I think there's this world where. Physical hands-on trades like that are going to be a thing that a, are rare because the training takes years to get there and we're not gonna get there that fast. And so I think there's gonna be this arbitrage, but then to have tools that are available that they've never had access to before, like plumber making their own website would is crazy. a, solopreneur plumber is not going to shell out the $50,000 that it needs to make a great website. and so how can we put. website for that person, for their business that is necessary and needed in, front of 'em. It's, it's a good cause, but b, it's exciting, I think. and that's, how did you land on that as, how did you land on small business as the place you wanted to tackle? You talked about it a little bit, like how, like why that, and I'm sure you thought about, hey, let's go after mid-market, or let's go after enterprise, or let's be the backend to Claude's, Claude system and create the new un browse and headless browser or whatever it might be.

Raj Singh

We had something there too. Yeah. Yeah. I'll say,

Patrick Patterson

small business?

Raj Singh

so our, so I came to Mozilla, they acquired my last company, and, without going the whole backstory of that integration and whatnot, I was moved into a team to build something new. And, at the time, this is pre gen ai, Right? but I knew there was this emergence of these Like, models and I knew that only because my last company was doing meeting summarization, pre LM era, right? So I've been in the AI category for a long time. and but we, had to do it with, we had our own team annotating transcripts. We had to do our own training. It was a d different era, right? so I saw, how good the tax generation was coming out of, at the time GBT two five, right? Not GBT three, right? And it was the thing I tell people, they're still winners, right? thousands of things being released. There are. Products that are winning now, maybe the economics have changed. Maybe these products aren't meant to be billion dollar businesses, which has been this venture model, like for the venture, for the early stage venture industry. It's tough. Like how do you pick in this market, only, there was no cha, GBT. And I had come up with the idea of building a website builder. and the reason was it was a combination of three things. One, I had this thesis, rule. I called the 95% rule. So assume AI is 95% accurate. And What sort of applications do best where you can be 95% accurate? so for example, if I'm doing product recommendations, 95% is fine, right? Because you can go and review it. If I'm doing content recommendations, 95% is fine. If I'm doing a self-driving car, 95% isn't good enough. if you're wrong, 5% of the time I'm gonna die. so that kind of check that box like, oh, website, oh, okay, I only be 95% accurate. And then human can go and further edit it and personalize it the way that they want, right? So that was one, the second thing I realized was, I was looking at Yahoo Finance and this like chart emergence of the fastest growing business in the US was the solepreneur, the service provider, the small business? owner. And this was really the outcome of COVID, and remote work. But you had the mural painter doing the side hustle, the side hustle economy. The gate economy, right? So I was like, oh, wow, That's interesting, right? So I'm intersecting those two things. And then I was also looking for a point of disruption. And I realized that the entire website industry charges for connecting a custom domain. and I was like, wow, this is like a Robinhood type opportunity to disrupt Rob, brokerages charge for trading fees because they used to have a physical stock worker do it. They moved to electronic trading, but they still charge for trade until Robinhood came around. So they realized this is stupid. this is an arbitrage, your, margin is my, opportunity, right? And so I was like, what if we just allow connecting your custom to domain and making it free, right? Nobody had really ever done that before. And I went and started digging into it and realized SSL certificates are now free. Those used to cost money too. And that was actually some work out of Mozilla that created this sort of. the free SSL, I'm blanking on the name of the service, but free SSL certificates, whatnot. And then I also realized like the cost of bandwidth has basically gone to zero. 'cause most small business websites get almost no traffic, right? So it's it's almost negligible, right? And so it's wow, there's really no cost to running a website. This is all like manufactured costs, right? And so I took those sort of three angles and I was like, oh, here's an insertion point. We can marry, technology defensibility by marrying g PT two, five, with this macro trench. So swimming downstream with, solopreneurs, small business owners. And then we have a angle of attack. 'cause we have a business model disruption, right? And so then there was born solo, which was an AI website builder before Chacha, but Tierra, right? and so it started growing in a viral kind of way. And the thing is distribution is what's hard. As we talked about earlier. So then we said, hey, what if we leverage, the hundreds of leads of pup a day, right? All these people, small business owners. And then can we translate that into upselling more valuable products to generate more revenue? And it, so it has become this entire small business suite initiative, within, Mozilla. So that's how got.

Patrick Patterson

that's a masterclass of how to see, see the opportunity, combine it together and attack it. And I think, in a world you mentioned, and it's, easy to create. hard to know what to create and, that time and thinking through it and seeing the trends and, and, then, figuring out what to create, it's, think it's a good lesson for anyone listening. you can relate that to, silly app that you're building on the side. can relate that to your company. You can relate that to that idea that you have to be a solopreneur on your spare, in a gig economy. I think it's really, smart.

Myles Biggs

So Raj, if you could go back in time, now that we've, we know your arc more and we've heard more about who you are and what you've done, all the way back to your side hustle of installing ethernet core cards to you're doing at Mozilla. If you could go back and tell your college age self, say, 25-year-old Raj, a piece of advice besides PRPs, favorite adage of buy apple. 'cause we'd all say that. what would you say? Because there's people listening to this that are that age, that are you at 25 now that have a chance to, learn from your story?

Raj Singh

I don't have a great answer for you. It's tough. I think, and the reason is I'm a big sort of subscriber to chaos theory. So we could have changed a decision at some point, and we don't know what the tenure downstream outcome might have looked like, even if I think it may have been a, let's say a materially better result. I will say reflecting, there were certain times I probably should have taken more risk, than I took. there was times, I probably should have leaned in, this sort of hard thing about hard things. My own intuition and decision making. but yeah, a lot of energy is spent getting teams on board, getting everyone on board, so move faster. counterintuitively could have said, Hey, just go take a job and stay there for 20 years. That's crazy. As that sounds, that's a completely different, path, right? it would be a much lower stress lifestyle. but, I think, the one thing I'd say that has been constant and hasn't really been something I would change, but it's probably the most common question I get, is like, when's the right time? And I'm like, it's always the right time. And we talked about that a little bit earlier. so that's definitely something I, advise people when they ask. I'm like, just do it. if that it isn't gonna go away, assuming you can afford the opportunity cost, and pull it off, give it a shot because there's no better time.

Myles Biggs

I love that. all right, my final question then, 'cause we have covered so much and you've seen the industry change in so many ways, what do you think is gonna be true say a year from now? not true today. If you had to make a prediction or hot take timestamp and send you in a time capsule this time next year, right or wrong, what do you think it'll be?

Raj Singh

this is a weird era. I think in tech, tech historically has always been a bullish market, like a bull market. Like it's always growing. We've been eating everyone else's market share. it's, a tough time for a lot of period, a lot of people, a lot of anxiety. the job market has changed, for a lot of folks in tech. I, have worked through a variety of crashes and you could see the sort of, light on the other side. there, I think for a couple years we were going through a little bit of AI skepticism and Ai ome doism, but I think now we're at a point where people are recognizing that workflows are changing. And I do feel by the end of this year, we should expect most enterprises to get to 60, 70, 80% sort of clot code or equivalent adoption. not superficial, just ask a question, get an answer, but it as a meaningful part of your workflow, in an agent way. and I think people are gonna struggle who aren't able to make that sort of transition, or at least invest the time to try to make that transition. Because, the, one of the hardest things that I've been thinking through and trying to figure out is what does the modern interview process look like? is it, how do you evaluate a pm do you still use the sort of PM craft and the 70, 80 competencies that comprise that or, I've started thinking about builders and specialists, right? And, just eliminating job titles altogether. So I think this is just all very interesting time. I think this year, we're gonna, we're gonna see more of this push around AI adoption and the pretty, I think at the same time, enterprises have to be? thinking a lot about AI governance, shadow it, security, SRE Cloud Infra, as well as cost. so I think all of those initiatives are running in parallel.

Patrick Patterson

Yeah, I think, I think it's gonna be, you talk about folks that aren't willing to make and I think the, argument I get, I, talk about the industrial revolution and, the automation of jobs and workers and it's that, they all got retrained and they did this and they went to school and all that. And, yeah, but it also happened over a 42 year period or something. then, I think we're in a world where, that is gonna be collapsed. I don't know what the timeframe's going to be, but it's going to be collapsed. It's not gonna be 42 years. I think we can all agree on that. or 70 years, or however long it take took to, to fully, get rid of the buggy whip. But it's, The advice Myles, talking about going back, like going to people today, the vi advice today of thinking about this future that Raja is talking about. It's be curious and, willing to learn and be willing to try it. when you hear someone like, like Raj say, talk about 60% Claude Code adoption. I bet you there's folks that are listening to that. It's yeah, I use Claude. It's the, that's the same thing. I use Chachi pt. That's the same thing. It's no, it's a completely different thing. and it'll get easier, and, better to use as, the days go forward. But be curious on what that means and why is that important and why are, enterprises adopting it at a 40% rate when it was 0%, like a year ago? these are, important questions you should be asking, and how can you adopt it in your workflows? So the curiosity, the willingness to, learn to fail, I think is really important right now. and I think that's, gonna be what gets people through over the next year. to two years. this was awesome. I don't get an opportunity for an hour to, to sit and jam with someone who is building really important products, like what you're building and who's gone through as many exits. I really appreciative of you coming on the podcast here. If people wanna get in touch with you, if people wanna reach out to you, what's the best way folks to do that?

Raj Singh

Yeah, I'm on, Twitter Threads, mobile garage, LinkedIn. you can DM me, Tre, whatever generally online. Easy to find me.

Patrick Patterson

Nice. throw, the, the contact information in the show notes here. thank you again. Really appreciate it. Myles. Take it away.

Myles Biggs

Yeah. echo it. Thank you again, Raj, and thank you to all of you for listening. Be sure to check out the show notes for those links that we just said and be sure to subscribe to this podcast for more good enough visit. We'll talk to you next time.