In this episode of The Next Next, host Jason Jacobs interviews Andrew Lau, co-founder and CEO of Jellyfish. They discuss Jellyfish's role in providing data-driven insights to engineering leaders, aligning team efforts with business objectives, and enhancing visibility into operations and productivity. Andrew shares his experience co-founding Jellyfish in 2017 and how they've successfully raised venture capital, including a notable $71 million Series C round. The conversation also delves into balancing the challenges of running a scaling startup while raising a family, the evolution and impact of AI on startups, and the shifts seen in the landscape from small teams to larger companies. Andrew emphasizes the importance of maintaining a focus on the value of team and customer impact. The episode concludes with Andrew offering insights into how Jellyfish is navigating and leveraging AI both internally and for its customers, and the importance of adaptability in a rapidly changing tech environment.
Navigating Startups, AI, and Family Life with Andrew Lau of Jellyfish
In this episode of The Next Next, host Jason Jacobs interviews Andrew Lau, co-founder and CEO of Jellyfish. Jellyfish is a Boston-based software company that provides engineering leaders with data-driven insights to align their team's work with broader business objectives. Andrew discusses the founding and growth of Jellyfish, including their successful Series C round led by Accel in early 2022. The conversation covers the challenges of balancing startup growth with family life, the transitional impact of AI on engineering productivity, and the future of AI in business operations. Andrew shares insights on navigating market changes, the importance of strategic decision-making, and the value of building a supportive team culture.
00:00 Introduction to The Next Next with Andrew Lau
00:17 Meet Andrew Lau: Co-Founder and CEO of Jellyfish
00:35 Jellyfish's Mission and Venture Success
01:03 Andrew's Perspective on AI and Business
01:58 Introduction to The Next Next: A Learning Journey
02:38 Reconnecting with Andrew Lau
03:38 The Early Days of Runkeeper and Jellyfish
04:22 Building Jellyfish: Challenges and Successes
06:58 Balancing Family and Startup Life
09:05 Navigating Business Phases and Trade-offs
21:58 The Impact of COVID-19 on Jellyfish
25:14 Focusing on Engineering Leaders
26:18 Shifting to Larger Companies
26:29 Mid-Market Sales Strategy
29:38 Adapting to AI in Customer Operations
36:48 Internal Use of AI Tools
37:01 Future of Teams and AI Integration
40:13 Disruption and Market Adaptation
41:06 Reflections on Startup Evolution
45:23 Jellyfish's Future Vision
47:33 Closing Remarks and Contact Information
Jason Jacobs: On today's episode of The Next Next, our guest is Andrew Lau, co-founder and CEO of Jellyfish. Jellyfish is a Boston software company that provides engineering leaders with data-driven insights to align their team's work with broader business objectives, enhancing visibility into engineering operations, and productivity.
Now, I first met Andrew I think back in 2007, 2008, when I was just getting Runkeeper off the ground. So he's a fellow fogey founder like me, and he's in an interesting seat at Jellyfish in that he founded the company back in 2017. They've raised a boatload of venture. They did their series C back in early 22, a 71 million round, led by Accel with Insight Partners and Tiger Global.
And they also are working with engineering organizations to help them run better and to help better map them to the business objectives of the company, which means that Andrew has the [00:01:00] perspective of building a venture-backed company while also raising a family. He has the perspective of seeing how AI is transforming how startups get built and funded 'cause he's living that day to day and he is also seeing how much it is or isn't infiltrating these large organizations that are customers of his as well. And his perspective's a really interesting one where he's a bull long term, but in the short term, maybe there's some overhype and maybe people underestimate, not just the technical implications and possibilities are one part, but then. Sometimes people underestimate how long things take get to get adopted because of switching costs, behavior patterns, existing teams, existing processes, legacy, infrastructure, technical debt, et cetera, et cetera, et cetera.
At any rate, fascinating discussion. I learned a lot and I hope you do as well. But before we get started.
I'm Jason Jacobs, and this [00:02:00] is The Next Next. It's not really a show, it's more of a learning journey to explore how founders can build ambitious companies while being present for family and not compromising flexibility and control, and also how emerging AI tools can assist with that. Each week we bring on guests who are at the tip of the spear on redefining how ambitious companies get built, and selfishly the goal is for this to help me better understand how to do that myself.
While bringing all of you along for the ride, not sure where this is gonna go, but it's gonna be fun.
Okay. Andrew Lau, welcome to the show.
Andrew Lau: Jason, wonderful to see you. I was actually counting the dates and I think it's been like 15 years or something. I don't know. I'm making up numbers, but it's been a long while since I saw you, so this is great.
Jason Jacobs: it has, yeah, it's it's bel we were just talking before we hit record, but it's just unbelievable how the years, fly by and I, my theory is [00:03:00] that is that it's like a nu numerator and denominator thing. Like one, the one year unit stays constant, but the denominator gets bigger and bigger.
Meaning it's a small and smaller percentage of your life, which I think is why it makes it feel like the, like the time goes faster over time.
Andrew Lau: Yeah. And I also would say the last five years have been so crisis to crisis that it has actually gone really fast and slow. Like in some ways time has moved on really fast, but then you look back, you're like, whoa. It's, it's like, it's a, because we're just, we're so in it all the time just dealing with this issue or that issue that it's been, that's been fast forwarding on this.
Anyways, wonderful to see you. Sorry it's been so long, but my memory is, I remember sitting down with you. Early mid runkeeper journey and like you were, I was like, you're like a recruiter, but you like ran and you're like doing this thing. And it was like the go, just make consumer stuff and magic will happen days.
And I don't know, I made rose colored glasses, but those sound like the ideal times back in the day.
Jason Jacobs: Yeah, no, that I used to be a pretty eager guy. It's like a, like old bull, young bull now. The, I'm [00:04:00] still pretty ambitious, but I don't know. My, I don't know, I don't know if my tails wagging quite as much as it once was.
Andrew Lau: Hey I always say this to everyone. Take the wins, man. So belated congrats on all the things you've done since we last talked. We clearly can't even hash it out in, in, in the quick before Chad nor even in this session around replaying all of it. But congrats, man. It's been cool watching you grow and change and evolve and do all the things.
Jason Jacobs: Likewise. And and that's a good segue 'cause I was about to congratulate you for what you've built with jellyfish, but what's jellyfish.
Andrew Lau: Thanks Jason. So Jellyfish now is actually an 8-year-old company. We help companies drive efficiency and productivity, but also help around software engineering teams. But really it's about helping teams get unstuck, making sure they focus on the things they actually need to focus on to make the business succeed.
I remember we actually started the company, we made a trite, glib remark, which is we were trying to do the sales force for engineering. It's been eight years, which is crazy. We have north of a couple hundred people on the team 500 customers. And the best part for me is I [00:05:00] got to found it in a work with Dave and Phil, who I met 25 years ago who hired me.
And
Jason Jacobs: As an intern, right? Weren't you an
Andrew Lau: Yeah. Yeah. It's been it's tremendous to get a chance to go do that. And actually, I'll say it for you as we get older, as we get the chance to be intentional, as you clearly are on this endeavor being thoughtful on why you do a thing and who you do it with, and all these things are really important.
And I got the chance in 16 to reconnect with these. We kept in touch socially, but to connect professionally and decide to do something together in 2017. Which is a privilege to get to do that. But it's, it is full of intention to want to go to battle with these guys, go to work with these guys, go to build with these guys is very lucky in that part.
And we were conscious in doing that, so be intentional.
Jason Jacobs: There's a few interesting threads which led me to, reach out Andrew. One, one is, it's just that we're of a similar age, and I know you, you also are right in the thick of it with kids and you're running a scaling startup with over 200 employees that's raised a bunch of venture [00:06:00] and and yeah, so one part is I just want to know.
How that is. 'cause
Andrew Lau: It's so easy, man. No,
Jason Jacobs: I can't imagine that right now. But then there's another piece which is that you wrote a piece, I think it was for Fortune, just talking about, from what I can gather from what you've said publicly, and we haven't talked a lot privately yet, which is why I wanna have you on the show.
But you're both. Bullish on the long term with ai, but also it seems like the data that you're seeing from the trenches is that in the short term there's some overhype. And given that you have the data on how it's actually getting used. In the field. I think that's an interesting thread.
And then I think the last piece is just given that you're bullish long term and and cautious short term just how that manifests in terms of how much you turn up the crank in terms of how you're utilizing internally and how it factors into your strategy. So we don't have to tackle all that right up front, but those are the reasons why I invited you on the show.
Andrew Lau: I am I'm down to cover all those things and I'll let you lead the way. 'cause there's lots to talk about in all those dimensions and I'll let you pace it. Or how you wanna tackle these things.
Jason Jacobs: For starters, just, you [00:07:00] founded the company in 16, your kids are a lot older now. Talk, talk a bit about kind of the phase of life you're in then, and the
Andrew Lau: Yeah.
Jason Jacobs: life you're in now and also just as the company's evolved, just what it's like to be a founder, CEO, navigating through all of that.
Andrew Lau: Yeah. Okay my girls are 10 and 12 now, and so I guess we started eight years ago. So they were two and four, right? At that time and maybe even when we started chatting with co-founders, like one in three. Look, clearly, really aware of the trade-offs and difficulties of actually.
Doing it all together. I think a younger version of me would've been chockfull of, hubris and be like, ah, startups, parenting hard, impossible. And look I actually, I don't fully believe that statement clearly. But a younger me would've actually said stuff like that. And I think actually it's that nothing's impossible.
You just have to make trades, I think is the part of it. And I think we were very conscious of this. 'cause when we started, all three of us actually had younger kids. I think Phil's kids were a little bit older, but [00:08:00] Dave and I's kids were about the same. And I think we actually said as a because we decided to want to do something together before we even came across jellyfish and this engineering management space.
As a concept we said let's try to do something together. But we said we have to make this work. And it was like a precondition, like we are all parents right now. We have lives on these things. We have families that we have to get there for a daycare pickup.
We have to help cook dinner. We have to be home to support. And of course we have partners that have been very supportive too, but also like we just have to make it work.
Jason Jacobs: Does that work in a scaling startup? That's raised a bunch. And I guess I'll ask the, I'll ask the question. Like what about peacetime versus wartime? 'cause you don't control, you don't control when the company's in wartime. And I'm not suggesting you're in wartime, but
Andrew Lau: No. No I actually like, yeah, we're more time, I actually think they come in surges actually. So I actually want to tackle this and look, I, I can retroactively look at it for us and I can talk about ways to think about it, but I can't tell you how your business is and nor can I tell you how this business is gonna be.[00:09:00]
A year from now. I think you just, I think you make your best choices and thoughts and you react to the situation. I would say look though, I would separate the different phases of the business, right? I think, like I was intellectually talking about the ideation stage. I'm talking about the zero to one stage.
I'm talking about, to your point, scaling now on this stuff. And they're all really different, right? I, I'm just saying at the ideation stage we just said, Hey, like we are. Parents and family people at this point. There's no going back on that, right? But yet we want to do a startup together.
So that was a reality of making the condition worked. And to your point, like I, my only, the point I'm trying to say is I'm not saying it's impossible, nor am I saying it's like, of course it's gonna work. You just have to make it work. And so for us it was like acknowledging look, B2B software sometimes has pacing that actually supports because sales cycles are longer.
You can rearrange time a little bit. Versus like you were living consumer stuff, which is, it just surges in these like haphazard moments on these things. And so I, I'm just saying that there are certain kind of [00:10:00] businesses that are, might be more conducive to reshifting time on this stuff.
But to also and look, acknowledging it's hard enough to get any company off the ground. Adding more conditions is difficult. But also acknowledging like how your team and split was like, so we said, Hey. As co-founders, we're all in the situation, but also like we just have to acknowledge it.
So this means that hey, we may not be able to, if do daycare pickup, you can't, be, like at the office, like hammering from six to nine and then hitting the bars right afterwards, right? That like that, like maybe that was the early twenties thing. That's not gonna work.
And so acknowledging that's not gonna work means like you make trade-offs. Say what's the team you're gonna actually hire? Are they going to be able to make progress without someone like sitting right next to them all the time? And acknowledging that hey, while we have to go for daycare pickup and actually get dinner made, we can be back actually at the keyboard at nine and doing things.
And it just have a more regimented schedule, right? So there's gonna have to be a. Work [00:11:00] smarter, not just harder thing. And so look, these are all, they may sound trite, but like we, these things assemble together to being a plan. And I think step one, just being acknowledged that part of it.
Now, a couple other things I would say that getting that zero to one or getting that first thing off the ground is hard, right? Because I think like all of that is very, all of this whole journey is emotional, but that's very emotional because you're sitting there trying to figure out like, can you get this actually off the ground?
And like sometimes you gotta will it and it's a different form of war time, but like it's till it's there, it's not there. And it's incredibly frustrating. And part of your brain goes again, if I could spend more hours here, could I do these things? Does it magically happen?
And I'm just saying that's a psychological thing that's actually hard. I think once you actually start having some customers and some business and team. You can start making logical trades being like, Hey, when urgency kicks on, like everyone around you knows how to actually respond when there's a customer issue or a fundraising or sailing, like new software.
And there's chances that can respond to the [00:12:00] system, but for zero to one you don't have that yet. And so it's really strength of your conviction and will to move it there. And I think that can be a very hard time for making the family trade offs and all of that part of it. And just like little hacks there, I think you can.
Be honest with your family and build structures on that stuff. Hey, I'm gonna need this. There are gonna be nights that I need to be there, so like maybe I can schedule it with the family if that's helpful. Hey, like I do, Tuesdays and Thursdays we're gonna do some networking stuff, or Tuesdays and Thursdays we're gonna go out with the team and just put it on the calendar so everybody knows it's happening.
But in exchange, I'll be there Monday nights and Wednesday nights to go do pickup on these things. Like I'm just saying these. Some, it's not perfect, but realizing that sometimes it's the variability that actually causes family tradeoffs versus the actual time. So just getting ahead of it, right?
So there's like little nets. And then I actually think at scale look, it's incredibly stressful, but I think things come in surges. And I think the good part is at a large scale you have a team. You can be clear around the challenges we all tackle together, but [00:13:00] there are times you've gotta dig in.
But I think one of the good parts of, a family that's watching your startup grow is they're empathetic and understand the structures of actually what's happening. And I'm lucky to have a supportive family that can do this, but like they kinda understand the pacing of the surges that we do need to dig in war time.
And there are times when other parts of the cycle like gives us a little more flexibility. And I think the biggest thing for all parties involved is putting it out there as much as you can. If you've got. Family commitments you gotta do is on the calendar forward so the rest of the team knows how to work around you on that stuff.
If you've got so your family knows that, hey, like this part of the quarter works out this way, this when board meeting works out, my kids know when the board meetings are happening, right? They get it right. And that's part of actually like just acknowledging that there are parts of the cycle that come and go and actually getting everybody ready for that stuff.
Does that make sense at all?
Jason Jacobs: It does. Yeah. And I think some of it is just I know if I think about my wife and I the stressful time is not actually one something. Is in motion. It's not actually the [00:14:00] thing that's going into motion. It's the uncertainty around before it's in motion and then all the what ifs. But then once it gets in motion, you just build systems around it and you adapt.
It's the same thing with if a family member gets sick or something like it, you don't wish it on anyone, but you adapt, right? And you just take the new normals as they come at you. And I think that's what you're saying. Is that it? Yeah. Yeah, it's stressful, but if you have systems and structures and plans and set expectations et cetera, then you adapt.
And it doesn't mean you won't have your moments, but but like anything you just build around the things that, that matter to you and you decide as a family, right? What, what matters to you?
Andrew Lau: And maybe I'm being naive, but I'm just saying I, I don't want it the other way. I don't want the formula to be you can't do this if you're a family. No I exist in this way and we're gonna make it work. We're gonna make this company screaming successful. We're gonna make our family system work on this thing.
It's I just don't wanna believe that it's impossible on that part of it. So
Jason Jacobs: Yeah. And speaking of possible or impossible, I think we're gonna come back around later to talk about how AI will change how we work and live directionally. But [00:15:00] before we even talk about ai let's talk about jellyfish. So you said the Salesforce for engineering. What does that mean in practice?
Andrew Lau: Yeah. So we provide tools to help leadership managers and teams understand what people are working on, where they're stuck and where they need help. So this could mean everything from. Like understanding what percentage of your team or product is like working on, product A versus product B or bug fixes versus new features to like, Hey, why is this team like stuck in the forecast of when this thing's in a finish slow?
It's oh, it turns out. That this individual's not getting their pull requests actually supported, or this team's not getting the funding they actually need. Or there's some real breakdowns in the SDLC, the software development life cycle on this team. And so given that kind of data centric view and actually how the company can make decisions how to make sure it's most most effectively using its resources.
And so the Salesforce analogy is a terrible one for product engineering folks, but it works for business people. Because look, when I was growing up in the [00:16:00] nineties, professionally there, there was, no view in how sales works, and now we're metrics up the wazoo and companies actually invest more in sales because they understand.
It's oh, like we're stuck in the pipe. We're actually stuck in call one. Or it's actually, we're stuck in these conversions. And so by looking through Salesforce and the data there, the business from, the board to managers, to the team members can actually figure out hey. Where should we reallocate resources?
Where should we do things to actually make sure we're doing, how can we forecast a few quarters out? How do we know what's healthy and actually working there? And so we make that trite analog being like, Hey we aim to provide that same kind of visibility and systems to the engineering management function, which has largely been, like when I started it was two of us nerds in the corner.
But now it's actually the engine of companies, right? It is making the product that companies sell. It's 40% of opex and is at least understood of most of the functions in an organization. And so that's our ambition to, to provide that visibility and understanding and insight to that department.
Jason Jacobs: It's timely in that the discussion I'm recording after yours is [00:17:00] actually with the day ai founders who are building kind of an AI version of an of. CRM and I guess my question is in the CRM world, there's of course Salesforce as an incumbent. And I guess you could argue that you know that HubSpot is giving them a run for their money in, in, in some regards.
What did the landscape look like on this side of the world? And does it also, from an acronym standpoint, does it have an acronym, the category that you're in?
Andrew Lau: Yeah it's not quite yet. So the landscape is actually look. I think the software engineering space came from a, boutique creative where just making things. And so most of the innovation in the last 30 years has been actually for people making the code. And frankly, like the folks, the way they do it today is not the way I was trained.
They're much better at it now. They cut me outta the code base accordingly on that. No but I think that most of the tools have been. For the actual engineers, and almost nothing's actually talking about how it engages with the business. And yeah, like most of the tools that you hear about, which is, whether it's the, the source control things like GitHub or [00:18:00] the project management tools like Jira, like those are really built for the teams and almost nothing around how they tie to the business.
And so our lens of this is Hey, how do we integrate that? To the business of software engineering, how do we tie those all together? And so really thinking about how managers and leaders and the business tackle that to your question and acronym no, it has not gotten the quadrants of the waves yet on this stuff.
One answer is it might I don't know where it's gonna shake as, but they I think they're playing around with software engineering intelligence. I think they're playing around with the words productivity. But I think when that acronym comes, it means that like we're finally getting there as an industry.
It, if you look at the Salesforce. Analog. It probably took, Salesforce was built on the back of Siebel in the nineties. It probably was a, it wasn't probably until 20 10, 20 11 that we got defacto metrics that we ran by. So it was probably a 15 year journey in that evolution of that space.
Before it became of course you needed a CRM and now we're talking about Salesforce, HubSpot. Now you're talking about day. That's a three generation path. Seeable, that's a fourth generation in play there. So I, I think these things take a long, like a while to evolve and we're looking forward [00:19:00] to those acronyms coming forward in this space.
And by the way say hi to Christopher for me. He and I actually have kids the same age. I actually remember he and I really early on in certain we hung out with a couple other parents in a community center. In the winter, just 'cause there's nowhere to go with the kids.
I think the kids were like this small, like running around. So yeah, like he and I actually have parallel paths in that sense. So it's very cool that we're back to back here.
Jason Jacobs: A, as a founder if there's an acronym for the category that you're entering, is it too late?
Andrew Lau: No, I think your playbook's just different, right? I think it depends what your play like. Whether you like greenfield spaces or you like, or what are they, the blue ocean or red ocean kind of things. It depends what you're looking for. I think if you're, if there's acronyms in play, then you're probably walking into a dog fight situation.
It's probably a fast follower game. It probably means there's gonna be limited product differentiation and it's gonna be. More about how you attack the market and go fast there. I think in pre acronym worlds I think you're defining the space and the struggles there are much [00:20:00] more making sure that people care.
You are actually they don't show up with budgets and you're actually elbowing out and probably replacing hand driven processes. So it gives you room for creativity, but like they may not have time for you on that space. And maybe people don't care, right? So I think they're just different fights.
It depends what fight you want to go into and what you're suited for.
Jason Jacobs: When you think about historically, so it's been, call it nine years or so since you started, if you were writing the jellyfish book what do you think the key chapters would be thus far in the story?
Andrew Lau: Oh man. I think like it, it'd be an exciting book. I like, so one, I don't think we're ready to write a book because I, as founder journeys are, you're always paranoid around what's coming around the corner. I don't think you get to go do that till the later chapters are written so paranoid of that.
I think there's so many chapters of this stuff. So one could be I think I would like, look, it's been such a crazy last five years, I think I would do it around chapter one [00:21:00] is getting the band back together. Chapter two is and deciding what to go do. I think the first couple years were just like, that early crew of the first, 10, 15 of us, 20 of us just like making something outta nothing.
And then and you
Jason Jacobs: And that time was spent. Was it spent building and experimenting or was it spent under like market research and development
Andrew Lau: I, I think it was about building or quote alphas with our first kind of 20, customers aren't even the right word, but Alpha Partners that're willing to work with us to develop stuff to figure out if they give a crap about we're actually making and I think all those people.
Willing to do it with us and all the interviews we did, but yeah, the first couple years are just making something out and I would actually argue that it could be incredibly frustrating for people. You want that to run, you read blogs all day about how these companies are taking off and yours isn't yet.
And it doesn't matter if you will, it it's it's, and you got the pressure of the team and investors and all that stuff, and you're trying to make something happen and then. And then I think I think you could [00:22:00] actually write it around the phases of the crazy world Covid hits.
And and then like we get that trends, that the movement to work from home and hybrid suddenly catalyzed us to be relevant in that moment. You could also
Jason Jacobs: what, and what was it about that moment that
Andrew Lau: so yeah. Look we. 2019, we're already starting to sell, but these are all a lot of small numbers, first 20 customers kind of stuff.
And then you get into Covid and I think Q3, 2020, I think we did more in one month than all of 2019. And the product hadn't magically changed it or the, or really the marketing bones. It was really the, it was the world changed. Ev everybody went from oh, I know what everybody's doing.
To, all of a sudden everyone's going oh my God, I don't see, I can't walk the floor anymore. I can't see anybody. Are folks okay? Are they happier? Are they working faster? Are they working slower? I mean there was that traits, are people working two jobs? They're working zero jobs.
Like it was all, everyone's I don't know what's going on. And people suddenly needed. Visibility [00:23:00] into what's happening. Just basic stuff. You got from, used to be like at the water cooler, you just didn't have it anymore. And so that was like a, that was a catalyst moment. And then I think then there was the crazy go-go 2021, like the hyperbole of that time period.
And that was crazy. And then you've got the challenges of kind of the tech recession of 23, which actually I. Meant that companies suddenly cared about efficiency and productivity. That because they needed their companies to survive and thrive in a scarce budget world. And that meant that relevance in that space.
And so that's another transition for the company and business and these things. And then we've got transitions again right now that we're starting to see around like this shift to AI and all of these things. And then you've got like the tumult of the financial markets along the way.
You surprise S-V-B-I-A funny family story. I've got a, my 10-year-old strangely asks me about the SVB stuff all the time. Why? Because I don't think she fully processed it all, but she [00:24:00] remembers we're sitting there in and COVID time and I'm like in like hair on fire resolving the crisis of oh my God, the bank's crashing, right?
And then she like. Knows it. 'cause it was just like all of a sudden I was just, we're all around because Covid kinda worked from home time, but it was like around in a weird crisis way. And so that's like a, I really don't think she knows what it means, but she asks about a SVB like thing randomly all the time.
'cause it's a moniker that's seared in her brain in that space, which is weird kind of parental crisis handling stuff right. On that stuff. So I think these are all like, I think it would be like. I don't think we're ready to actually talk like to write the book, but to your question, I think there's like all of these interesting chapters that appear that have been like, really dynamic.
And like the world's been going lock to lock constantly on these things. I think it's been like, I think it's been hard for everybody.
Jason Jacobs: If you, over that time, so since kind of 19 or 20, when you were serving customers in earnest to today [00:25:00] how has. Who you've served evolved, how has what you're doing for them evolved and how has how you go to market evolved during,
Andrew Lau: Oh man. Okay. Geez, these are good questions, man.
Jason Jacobs: I'm just curious. That's why I do the show. It's like these are things I really wanna learn
Andrew Lau: Look we, early on, like we started this, which is we said we knew we wanted to focus on engineering leaders. 'cause we had gone through the journey ourselves. All three of us grew up as engineering and product leaders. We knew the pains.
A lot of our community was in that area. So we originally like many startups it's your, you even used the word earlier. We were talking around founder market fit, like we probably had that kind of in market intelligence. We had lived in that space. So a lot of our network were those people.
So our early customer development calls or early alpha partners were people that, we connected with and we knew. And I think like early on, like even we're milling around in in, in eight, 17, 18 we thought we probably were working more towards smaller customers. And so eight round startups and stuff like that again.
And we quickly learned through that journey that like, look, when you got a team of 10 people around the [00:26:00] table, you know what everybody's doing. And not only that, you're not under kind of zero sum financial pressure around making trade offs. You're not getting ridden for efficiency.
You're getting ridden to be relevant. And while we add value there, that wasn't gonna be a, like a good like way that we were going to. Like really provide value and kind of innovate in that space. And so we quickly realized we need to go to larger companies, and that meant kind of BC round kind of companies.
So that kind of shaped things. So think north of 40 engineers kind of stuff like that. I'm hand waving here. I think from a go-to-market perspective we've always known we actually wanted a kind of mid-market motion. And what we mean by that is like this is a founder comment.
Look, when we were growing up, it used to be if you're selling a hundred south of a hundred K deals, the math wasn't gonna work to have professional sellers. And so you could have like credit card things like sub 10 k things if we were swiping credit cards. And then you could have north of a hundred k 10 sellers in the field stuff.
And so we actually decided, we actually wanted neither because we felt that the value proposition we're innovating for was not gonna be a credit card swipe. It's gonna be teams at [00:27:00] scale. Which meant, going through. A sometimes more complicated sales cycles, which means you just couldn't do it on a credit card.
I can't swipe a credit card for, for north of 10 k kinda stuff. And then maybe you can but then the, we also felt like we didn't want to be doing really big enterprise sales from the beginning. 'cause we'd lived the life before. When you're selling these. High six, seven figure deals.
Not that it's easy, but if you do the sales cycle, take, enormous time, periods, years, on the stuff, but all that at those dollar values, you're building a lot of custom stuff. For buyers. And from a product perspective, it is possible that you degrade into a lot of custom work that's not repeatable.
And so we actually made a decision that we really wanted to be selling a mid-market price points which is, when I was growing these be called the Valley of Death 'cause you couldn't fund a sales guy for it. But I think post in this kind of Zoom generation, I think you can, and it turns out.
Our buyers, which are engineer leaders, don't always want to be, have people, visiting the office. And it actually worked in this covid time period
Jason Jacobs: You have a Salesforce, but it's an inside sales
Andrew Lau: say, inside Salesforce selling kind of [00:28:00] five, six figure kind of price points in that space. And that would've been that would've been that was novel in, in, in 17.
And because in 17 everybody was selling PLG bottoms up self-service things and we're a little contrarian and we just felt like that. Our value proposition, the place we wanted to innovate for was not well serviced by self-service things. We needed to evangelize. We needed to get out there and meet folks that weren't just like out there shopping online they needed to be sold to in that space and talked to and reached out and listened to.
And so that meant that it was going to be a top down selling motion, but a remote one on that space. And so anyways we willed that existence on there. We wielded it by bringing in BDR SDRs first I guys to help us call into those spaces. And a lot of early founder selling and then supplementing that with sellers once we figured out some motion stuff.
And then the last part of the journey is I think in the last few years we've been, going up market in the space and we're now privileged to get to work with public software companies in that space where the deals get to be bigger on that space. [00:29:00] And that we actually have.
A larger plethora of constituency on the buying side. But I do think though, as a founder, it doesn't mean it's gonna work. You have to choose where you wanna play. Both in, where the value proposition, who you wanna listen to and what price point. It doesn't mean it's gonna work, right?
But you have to be somewhat intentional because the market can pull you and investors and team members can yank you in one way or the other. And you may choose that you don't want to go there. Doesn't mean where you want to go, will work. But I think you have to have an opinion because I think motion and how you listen to does have hand to hand influence on product and go to market.
They don't live in a vacuum. They go back and forth in that stuff.
Jason Jacobs: So when it comes to ai, when did you first start seeping or seeing AI seep into how your customers were?
Andrew Lau: Oh yeah.
Jason Jacobs: utilizing and how's that evolved? And then same question for your internal operations, either on the product side or go to market or both.
Andrew Lau: Oh, so many layers here, man. So look I think we started, look, we have all been reading [00:30:00] about all the stuff in AI for a couple years now and I use the word couple rounding in that space, but I think we started really seeing this in earnest from our customers in, gee, this would've been, let's see, this is.
Kind of mid 23, right? Mid, spring, like mid 23, our customers started saying Hey, we're really a bull on all these, all these coding assistance stuff. But we don't know if it's like we're getting the yield in our, for our ambition and for our yield, for all the dollars we have here, and we don't know why.
Can you help us figure this out? And so we're already working with them, said, sure. Let's. Let's do cohort A versus cohort B. They're using it and they're not, and let's juxtapose these things. And we got a chance to work with a few folks and we did a couple of talks to get around some insights around giving back to them around like, why, what was happening in this team from a metrics and change perspective, what was happening in this team.
And then we did some bespoke studies for a couple of them. We did a couple of talks and papers on that stuff. [00:31:00] And then we got a chance to work with the folks at like at GitHub and then they, we said, Hey we have a hard time kind of understanding who's using it.
And so they we worked with them and they got us an API to give us some telemetry of what people are doing, who's engaging with these tools. We don't actually have to cohort A, cohort B. And it's been really powerful for us to get a chance to, launched this feature.
So we launched it midyear last year and we now have around like 300 customers that are actually like engaged where we actually have day by day fidelity of the engineers engaging with these tools. And then and then actually furthermore, like then we, because working with them on the rest of the platform, we actually get to see the outcomes.
We get to see what's changing and the speed and what code is working, are they doing more roadmap or less bugs and what's actually changing behavioral, what kind of teams, what kind of projects. And it's a really neat corpus. And then now we've actually layered in. Other coding tools too.
And so we actually have this juxtaposition where you can actually see which kinds of tools are working, which kind of projects, what's actually happening, what the adoption curves look [00:32:00] like, and all of these things. And so it's a really neat and insightful dataset where it's not just usage, it's usage correlated with the projects that they're using and the organizational structures and teams.
And tools and then also then correlate with outcomes and what's actually happening there. So it's been a really neat study to actually watch it move through. I, I think kind of nine months of journey on that stuff. So yeah, that that has been the basis of why we've been able to make some really neat observations there.
And you mentioned earlier we've been actually able to publish a bit around that stuff. And there's more jewels to dig outta there. We're like, we're just still learning more as we actually do this,
Jason Jacobs: in a minute I wanna switch to how you're leveraging it internally, but before we do, I'd love to understand, just double click on. What you just said. So how much is it helping, is there consistency across the clients that you're observing and yeah, like what are the big, what are the big takeaways from what you've been seeing so far?
I.
Andrew Lau: Also I actually think there's like multiple layers. I think there's like different companies at different [00:33:00] sizes and scales. There's different kinds of tools being used and all of these things. And in some ways I think all the vignettes are true, right? So on one hand you hear from the valley around, the, these new startups that are actually getting a ton of leverage from these tools like Sier and stuff like that, that are like amazing.
Like I played with those that are really cool, right? Like I think Chris is gonna go off about this. He loves it. And I've talked to him about it and like. Where they are getting a chance to reshape their team and they're building this knit new code. And in some ways they're building their MVPs and prototypes and those early things really fast, right?
Because they can, they can work with these tools and orchestrate it in that way. And so I think there's some truth around that stuff that's actually happening. It's really neat, right? At the same time, we also get a chance to look at and work with companies that are at scale, that are going from be around a public that are, have hundreds and thousands of engineers implementing these tools, written mass.
And those companies are also seeing d different things too. Like they're seeing yield too, but it's actually like the micro vignettes on this stuff is they're not like working with half their team [00:34:00] sizes. They still have all their teams. They have not made trade off decisions on this stuff, but they're seeing, 10 to 20% lifts in, in any number of metrics you want to cut it by.
You could cut it by the, the cycle time, how things are actually moving. You could cut it by the throughput of prs. But in mass, in aggregate they are seeing lift, but it's on the order of kind of 10 to 20% on that stuff. And also you're seeing interesting adoption things where despite rolling this stuff out, they're actually getting only adoption of the kind of usage of 60, 60 something percent, two thirds of the team using it on a regular basis.
And you could probably even get more out of it. And so my takeaway on this very coarsely is that like we are on a transitional journey, right? So I'm a bull I'm using these tools myself and I think they're really powerful, right? At the same time, companies at scale are living with an existing code base they've already sold.
These tools are not as much accelerators in when they're not doing net new. I think there's value there because. So much of the engineering time is around, not just the coding, it's all the other [00:35:00] stuff they're doing. Some studies say that like hands on keyboard are only 10% of some people's time.
Furthermore, the work they're doing is around micro editing, nudging, changing more than net new. And then there's also this whole discussion, which is, it's just. Like these teams and organizations need training and enablement and adoption and goals and telemetry and measurement, which is where we're helping them with this stuff.
And the team is, that's using it doesn't really know how to use it and when, and so that's why the adoption rate to utility there. So despite it actually seeming like not a full yield set of numbers, I'm actually a bull even in that setting because. It's the unexciting stuff, training enablement goals.
Like you can just do those things and then you'll get more yield out of it. And so I'm actually really a bull for this whole thing because so much is changing so fast, right? And so like the innovative part of my brain is this is so freaking cool. And I'm also excited because the things we're looking at today in id, that's only like generation, inning two of this stuff.
There's next inning there's gonna be [00:36:00] all different kinds of tools, agentic systems, and all different parts of the SDLC and other parts of the day that we're talking about and how those things actually fit in. And how those things actually work. And then I'm excited also because when those appear, these companies at scale won't know how to actually use them.
Exactly. They need help with telemetry, understanding how to apply them, and that's where we can bring them value and then they'll start seeing some yields too. I'm excited for the transformation Our industry actually has to, because. The teams as they look today won't look like this in the future too.
And so we've gotta figure that out together. And so you gotta keep your hands close on those things. So I think it's a really exciting time for that part of it. And so yeah, I'm a bull, but I think that's like the juxtaposition of the telemetry and pragmatism of what I see today and what we need to do to get there in the future on this stuff.
Jason Jacobs: So what then are the implications in terms of how and how much you are choosing to push the envelope with these tools internally?
Andrew Lau: I like, look I would say [00:37:00] this for all organizations. I think the team of the future when you squint forward is I'm actually not a person who believes it's gonna be a hundred percent robots, nor am I a person that believes it's a hundred percent all humans in exactly the configuration we are today.
I think the team of the future, again, is gonna be half and half like AI stuff and humans on this stuff, and that means the interaction and role on that stuff. Is really gonna be different. I don't think any of us can like, like really know what that's gonna look like. And I'm just coming out from a first principle's perspective that I don't believe in either of the limit conditions right now.
And so I think it's somewhere in between. I think the journey of arriving there though, is gonna be through trial and error and telemetry. We need to be in there and actually seeing it. So like on our team, we're using all the tools, right? Like we actually, people say which flavor? We use all the flavors.
Why? Because. Our customers are using all the flavors, right? We have to try all the flavors to do it. Now we also are a software company, so we have to do this in a secure InfoSec like way. And so we actually have to make sure that we are doing it a [00:38:00] way that is that is a judicious with for our customers and that we're not.
Leaking data in this way. And so we actually have to also spend on actually making sure that we InfoSec and actually are hosting it appropriately or using the right versions of these things. And so we are actually balancing all of this stuff. We are, we are pushing the teams to constantly try all of the new flavors things when we can on this stuff because we want to blaze the trail for our customers and where the blazing isn't just the six person new school company.
It's also about blazing hey, we're a real company with real employees that has to make these trade ups. We're actually living the journey that our customers have to live more than like some new, four person startup. We like our customers, our companies at scale, and so we push our companies to our team to, to challenge the status quo of how to do this because we need to.
And that means figuring out like, Hey, how are we using these tools internally? How do we. Tie into these tools. How do we get yield and [00:39:00] leverage of this stuff? What's working, what's not for us? What's the telemetry to know that's working on, not us, let's dog food it. But then also like how do we provide differentiating features for our customers so that we can actually capture this moment?
How do we think about this critically in existence proof way? So like in a 50 50 AI way, like how do we exist in the future in that way? I think you have to keep asking those questions. At all three layers on a regular basis. 'cause the stuff is just changing so darn fast on this stuff. I think it'd be, I think if we were asleep at the wheel, I think it'll pass us all by,
Jason Jacobs: some of the companies I've had on the show are have strategies where they might have a SaaS model themselves, but they target companies with dozens of SaaS. Point tools that they're paying for. And through a combination of AI and open source, they enable consolidating into a single SaaS subscription and then ripping out in some cases, several if not dozens of tools.
Mark Andreessen famously wrote several years back about how software is eating the world. I guess my question [00:40:00] is AI an open source eating software?
Andrew Lau: Oh, interesting. You're making a markets question. Does AI
Jason Jacobs: It, and I guess you could take that a step further and say, do you worry about that as it relates to someone doing that in your category?
Andrew Lau: I I think look, you should, that's just general disruption. Are you gonna be disrupted? And I think you should be critically asking yourself that. 'cause if you're not, you're sleep of the wheel. I think that is founder Fear and paranoia. You should always be paranoid of that stuff. Look, we're still running, so I think we need to actually keep, we have to, try to disrupt ourselves and not innovate ourselves in that stuff. I, I think look open sourced, does it actually eat? I think that yes, I think there will be disruptions in point solutions in that way. I think we're privileged at a place where I don't believe that engineers will disappear entirely.
I think RD functions will still exist. I I'm a bull for that part of it. And so for us it's about playing into that arc of that transition for us. Because look, if I thought engineers wouldn't exist, yeah, we should be doing something different. I don't believe that. I think we will.
And I think we are here to actually manage through that transition. But I think you should critically look at that on a regular [00:41:00] basis. But I think there will be change, I think be naive to think that there won't be change on those things. But it's about where you play in the market. I'll actually I was on a panel recently and, someone asked Hey, with all these startups making all this AI stuff what does that mean for startups in general? And I actually think, I fell into, I thought about it for a second and I was like this was actually the same argument. If you remember in the web two phrase we're like, oh my God.
Look with web two and cloud software, you don't, you can start a startup for a couple hundred grand. And so like VCs and the whole ecosystem is gonna be upended. This is the end. What did you, you and I lived through that. What did we learn? We learned that no startups did not end.
It actually meant something different, which is actually, it meant that every startup could get going for a couple hundred grand to build a prototype. But the problem was. That meant that actually just generated inflation, which meant it needed more differentiation. It became really hard to differentiate against the next guy.
'cause it used to be like in the early OTs, if you made something, oh my God, that must've cost you 10 million bucks and now it costs you a [00:42:00] couple hundred K. It turns out everyone can make something. So you actually needed to make something better. You needed to market better. And it turned out, it actually then the talent war actually came about and you started need to raise 10 million again on that stuff.
So I actually think it's yes, every company in YC can actually do this with two people, not four. And they're building more stuff, but that means that all of them are running too. And so I actually think though, like the making might've gotten easier, but differentiation might've gotten harder.
And I actually proving value to customers got harder. So it actually means that you have to listen harder, right? To and as a founder, your role just is listen and retuning and shaping just got more important. So no, I don't think it's the end of vc. I don't think it's the end of startups. I don't think it's the end of these things.
But yeah, things accelerate, things shift, the ball moves. You have to provide value harder. You have to listen to your customers harder, you have to innovate harder. But the innovation may not be stuck in just the coding part of it. It's around iteration, refinement, listening, building that, getting to that product market fit part of [00:43:00] it.
I dunno, I just feel it's the same echo conversation we're having in that web two phrase. I dunno if that connects with you. But I that, that lit up in my brain when someone asked this question the other day.
Jason Jacobs: Yeah. No it does. I think it's just sorting through. It's okay, these changes will be profound, but then anytime it's this is killed, that's killed this thing's over that it's like inevitably it's almost always overblown and then the truth is somewhere. In the middle, one thing that makes me wonder though is if you were starting jellyfish today what do you think the implications would be in terms of the humans required and the funding required to get to where you are?
Would you, do you like, and I guess said another way, should growth investors be scared because companies can be so much more capital efficient.
Andrew Lau: Look I don't know, right? But I would say like inevitably, yeah, the playbook changes, right? Look. Back to my web two analogy. Like you start up then yes. You'd build on Heroku or you'd build an AWS and you'd actually be doing a different stack and like the developers would be different.
[00:44:00] So Sure. If you compare a company from early ts to the, 2010s, would they be different companies? Yeah, they'd be different companies, right? The way the who you hired and the sequencing is different. Inevitably though, to get to scale, you still needed the same amount of capital across both of those things.
So I, maybe the capital shifts a little bit. And look at all the companies today in the AI space, they're raising more. So maybe it's not even about less, it's more, so yeah, like I, I think things adapt. I think the skillsets change in who you hire. So it'd be naive to think that you'd hire the same people from 20 years ago.
And then on the flip side, like capital sequencing. Yeah, I think capital sequencing is gonna change and it'll change again in three years too, man. It always does. So yes, things change. Does it mean existential for parties? No, I think the markets adapt and like we still trudge on so yeah, like I look, I'm looking at you doing this building in, in, in the open and I think it is an exciting time.
So you get to look a little bit to the future and start differently, which is really cool. And as parents, you can say it too. I was just joking with someone like, look, my, my kids are 10 and 12. Like I think it's [00:45:00] okay. Like they can, they, we can see a little bit towards the horizon and be like, Hey the job's gonna be different.
So we get the chance as parents to tune them a little bit, but a little more of this, a little that not that we have any control with it, but they get to be aware of what they're actually playing into. So yeah, a little luxury and timing on that one. You can't control timing, that's what luck means in founder talk
Jason Jacobs: I know we're almost running up on time. To, I guess move towards. R wrapping up, when you think about the future if jellyfish is ex is successful to your wildest ambition
Andrew Lau: be like you. Yeah.
Jason Jacobs: yeah. Of a full-time podcast with no business model. But but yeah, what what does jellyfish look like at that time, and what have you achieved?
Andrew Lau: Look I hate, I don't like to put my brain in that space because it's, back to my earlier point, we always should be paranoid. That is the founder journey. And that means that we don't get to take victories along the way very well. But that is the founder journey. I think look I think I've said it before, but I think I care.
Most about is actually making a dent in the world. And I think for me, the dent [00:46:00] lands in two places. One, you make a dent in our customers and industry that like that, that, hey, what we did helped them do their jobs better and differently than they were doing before. And like that is a, and that what we do lives on.
And I think that is an important dent to make. And the other dent is in the lives of the team. Like I hope that the team gets to look back and go Hey, that was a magical time, and they all get to do things together in 20 years I was clear beneficiary of the diaspora of a company called Endeca Success that spawned off many companies like across the country on this that, that have gotten to amazing successes.
But also that I get to actually work with the same crew, the folks that hired me 25 years ago. That's freaking awesome. And I hope that I can make a dent in my team's festival. They get to have that conversation in 20 years too, in their own way. I hope they go crush it.
Jason Jacobs: I like that it's like a local versus, versus global. It's so many people's answers are like, I, we've solved global poverty or we've, [00:47:00] right. And it's like I I had a. I met with an old Runkeeper colleague the other day, and one of the things that she mentioned was just that the, like the tribe of people that worked there at that time, they're super tight, beyond any place she's ever worked since and that's really meaningful, right?
It's I mean it's on a financial stat. It's not a fun return or whatever, but it's it does mean something and it's it shouldn't be poo-pooed. It's hard to do.
Andrew Lau: Yeah look, if I am successful at those two things, then the other things actually happen to correlate, right? If you make a product dent in the world, you're generating some returns for somebody, like that. Simple as that, right?
Jason Jacobs: And then last, just, I'll just give you a chance for an advertisement, if there's anyone you want to hear from, either if you've got key roles you're hiring for, or certain customer profile or even just parting words you want to you wanna impart with listeners.
Andrew Lau: So most folks listening here, you, you probably are wishing for a bit more innovation and a bit more yield for your scarce budget from your software r and d team. Please talk and also folks that are trying
all these AI tools. And really into them, but maybe not seeing the full yield that you [00:48:00] wish you got outta it.
Let's trade notes. We got some tips we can help you get from here to there.
Jason Jacobs: Great. And how should people get in touch with you?
Andrew Lau: You can find me on LinkedIn. You can find me on X and you can mail me at am lau@jellyfish.co.
Jason Jacobs: Okay. Thanks Andrew. Thanks so much for coming on. I appreciate you making the time. I'll continue to be cheering you on making the Boston ecosystem proud and and yeah, let's keep in touch.
Andrew Lau: Word. Thank you so much for having me here, man. Take care.
Jason Jacobs: Thank you for tuning into The Next Next. If you enjoyed it, you can subscribe from your favorite podcast player in addition to the podcast. Which typically publishes weekly. There's also a weekly newsletter on Substack at the next next.substack.com. That's essentially for weekly accountability of the ground I'm covering, areas I'm tackling next, and where I could use some help as well.
And it's a great area to foster discussion and dialogue around the topics that we cover on the show. Thanks for tuning in. See you next [00:49:00] week.