Mark Barbir and Sanden Gocka, CoFounders @ Earmark

Welcome to another Founder Wisdom pod. We have the two co-founders of EarMark with us today, Mark Barbier and Sanden Gokar. Guys, I'm glad to have you on. Had a look at EarMark, looks pretty sharp. Maybe you can start with a round of intros, starting with you, Mark.

Yeah, hey Charles, thanks for having us. So my name is Mark Barbier, I'm the co-founder and CEO of EarMark. I've spent about 20 years building, scaling, and leading product engineering teams in B2B and B2C SaaS.

Charles, I'm Sanden. I'm the CTO of EarMark, and I've just been in love with engineering for the past, I don't know, 10, 13 years. Just been doing that ever since. We've done like 4 separate companies together, not necessarily as co-founders, but just kind of just working with each other on various teams. And And now we're here.

The app, when there's so many out there—

I'm sorry, Charles, you cut out for— you cut out really for a second.

Oh, I, I was asking why were you— did you guys decided to start a productivity company when there's so many out there? What's different about EarMark?

Yeah, so, so, uh, EarMark, uh, is a productivity suite where the work completes itself. So EarMark listens to your meetings and in real time turns what's said into finished work, things like docs, tickets, prototypes, updates, and next steps, unlike generic meeting tools. So product teams can move forward without all the manual follow-up. So it's not AI notes, it's not summaries, it's actual product and engineering work. For us, going to the productivity space, we really want to build a solution that we ourselves would use. Um, when we started the company originally, the premise was something that would be daily use and essential for people like us, right? We actually started with the Vision Pro solution 2 years ago, and then of course the Vision Pro didn't play out the way it was supposed to. We pivoted to a web-based solution. Now we're about 7 iterations in, and, you know, we've accomplished the goal, right, of daily, then essential to our enterprise customers, which is really exciting.

Interesting. And you guys decided to go bootstrap close to the Valley. That's pretty contrarian.

Yeah.

Yeah.

So our last company, Product Plan, which is a SaaS platform for product managers based in Santa Barbara, really close to here. Um, that company was sold in 2022 and that basically just gave us the ability to bootstrap and enabled us to create this business. Um, I think our goal is to get it to, you know, just, you know, some nascent PMF, early PMF., and then, and then, uh, potentially pursue funding from there.

Interesting. And ProductPlan itself, $10 million ARR is what I think was, is, uh, where it was at when you guys left, $50 mil acquisition. So you, you started that one from scratch. Can you tell me a bit more about that one and your lessons from, uh, your time at ProductPlan?

Yeah, so we weren't the actual founders. We were the first, uh, engineering leadership hired in a ProductPlan. Before that, we both ran software engineering at a scaled enterprise SaaS company called Mindbody. The things that we learned at both ProductPlan and Mindbody were, you know, in serving product teams and then managing and leading engineering product teams was, you know, every day, all day, you know, we're often in meetings end to end. Deep work is elusive. It's really challenging. Oftentimes what we experienced was the only time we could accomplish actual deliverables was early in the morning before people got on Slack. Late at night when people got— went home, and often on weekends. So the whole premise behind the service and what we've been trying to solve for is, can you basically just generate the work that you'd normally create at that time, like during the day? So the largest surface area for a lot of folks in R&D is meetings. That's where most context exists. That's where they spend most of their time. And then the idea of taking that context, leveraging it, and creating real deliverables from it. Shannon, do you have any color, anything you want to add?

Yeah, it's really like, you know, when I was a manager and, you know, having several direct reports, like it is, it can be challenging to essentially keep up with all of your one-on-ones and then also being, you know, expected to code and just kind of do all of the things. And so as Mark was saying, just to really like, can we create a tool essentially for ourselves just to like help us be a lot more productive? You know, in the day? Like, can it actually take a stab at like writing the Jira tickets for me? And not just writing it, but actually pushing it into Jira. Of course, having a human in the loop, you know, would be ideal, but just like examples, like examples like that.

Do you feel you've solved that with earmarks? Like, what's the specific bottleneck in that case that you would need to solve so that product people would actually have time to code?

Yeah, you know what's really interesting is we've— in terms of like the bottleneck, what we've noticed is especially with the tools like Codex and Cloud Code and like the amazing models like Fable that just came out, is it's so easy for, you know, engineers to essentially just create a just a huge slew of just code and PRs and like going off the backlog and What's not really being talked about too much is, you know, should all of those, you know, features actually be pushed into the product or is it creating kind of feature books? So like the bottleneck is actually kind of turning a little bit more on the product manager, just really needing to decide, okay, what should we ship or what shouldn't we ship? And so what we're trying to do is we're trying to enable a product manager essentially to keep up. It's like, what are these tools that,, you know, the engineers are getting, um, and can there be a tool specifically for either the PM or people on the delivery team, um, so that, uh, you know, they're not having to like wrangle with, okay, like, what is this engineer working on right now? Okay, great, I have that context, I could see it, maybe I, I know about it ahead of time, you know, before the meeting, because earmark can be proactive, um, and just like enabling them to flow, um, at that kind of same speed.

And you guys Yeah, okay, go ahead, Mark.

I was going to say, Charles, you know, the one way to think about it is like a cursor for product management, right? You think about engineers, right? They need an IDE. They're doing deep work in front of an IDE all day. That's where context exists. That's where they play. For us, we know that our ICP is in meetings all the time. That's so we're basically creating the same analog for them.

Just random questions, right? Like very technical people hate meetings. Why wouldn't they just all send their agents to one place and And I'm saying this because meetings usually are, is where emotions are discussed. And these folks, most of these folks don't care about emotions. They just care about like code and getting stuff done. Uh, or that might be naive perspective, but like, why they don't just send an agent and just be like, oh, I'm gonna keep my head buried in code because I hate meetings anyway.

Yeah, I think there's, there's definitely a profile, right? Um, but I think when you, when you think about scaled enterprise SaaS, um, the, the amount of collaboration and orchestration overhead, um, you're not going to be able to solve that with, you know, purely agent-based workflows, right? Like typically it's, you have horizontal, you know, dependencies, say your peer group, right? You're managing up to executives, you're managing down to individual contributors, right? You're talking to customers all day, every day. And the issue is, yes, I could send an agent to go do my customer interview or discovery, right? My customer probably is going to feel very good about that interaction in the way that my executives are not going to feel good about a bot joining a call and trying to automate communications to them, right? So the thing is, so I don't think that that reality will really, I mean, in some corners I'm sure it'll take hold, but we don't think that's the way it's gonna be in the future of work. In some ways, the way we think about the future is the idea that like meetings may have been a low-value activity forever, right? But in the future, it will be assumed that meetings will actually create deliverables, right? So the idea of me leaving a meeting when things aren't not finished will be like an anathema, right? Like the future of work is you're going to talk, things will offer themselves, and then things will get done.

And why just product people? Are you starting with that sub-niche? You're gonna get a very strong percentage of that niche, and eventually you're gonna go to other niches? Is that the play, or—

Yeah, it's like, it's like any classic startup, right? You're limited in resourcing, you have to have a wedge. So product managers are our wedge, specifically senior product managers who typically have multiple delivery teams. They have the most communication and coordination overhead in large— like, the larger the enterprise, the more complexity. We do better with more complexity. So that's our audience. And then what ends up happening is once they adopt it, you know, engineers, they get a whiff, right? It kind of scales to the delivery team, go-to-market teams. We have a lot of executives using it now. So even though we do have a wedge and we're trying to solve for product engineering jobs in a very specific, intentional way, it kind of takes a life of its own, which is really exciting.

Interesting. The GTM, what does it look like? Is it outreach? Is it events, organic traffic, ads? Like, how do you manage that part?

Yeah, it's classic founder-led selling through cold LinkedIn and cold email outbound. And basically, you know, what we'll do is we'll typically get, you know, a PM or two, right? See if they can kind of get a group evaluation going, get a director, get, you know, sort of the procurement purchasing people, get, you know, group evaluations going, and then have success criteria from those evaluations, you know, usually 30 days or so. And then that kind of gives us a red light, green light in terms of whether or not people will purchase the product.

And what's special about that ICP? How do they think differently from ICPs you've targeted before?

They do think differently. I think for a lot of product engineering leaders, there's a lot of FOMO out there around just their AI stack. Do they feel that they're going to be equipped for this future work? Do they feel like they're going to be left behind? So I think that's much more pronounced. For the audiences that we're serving, which makes sense, right? There's a lot of bad press out there, right? There's a lot of energy just around being relevant. So that's kind of why I think people are really receptive to an offering like ours. That said, the product just works too, which helps.

And the product role, it's somewhat new, right? But it's one of the hottest roles. How would you describe the history of the product manager and where's that role going? Could it be replaced by AIs and agents eventually?

Yeah, so product management as a discipline has been around for, I think, traditionally, like in current form, about 40 years, right? And it's kind of gone through a lot of evolutions. Today, the idea of a product manager is essentially the single sort of wringable neck, right, for a delivery team, a mini CEO. Or delivery team. And what they're in charge of is all the decisions around like an initiative and then strategic direction, customer empathy, right? Understanding what the customer needs, go-to-market, all the dependencies. They're usually servant leaders, right? Which means that like nobody reports to them. So it's a very challenging role. The reason why Sam and I started with, you know, one of the big reasons for starting with PMs was I think they actually have one of the hardest roles in modern work. They have the most diverse audiences, they have the most diverse deliverables. So that's sort of state of the union. I think for the future of PMs, we're very bullish. Dan Shipper just did a piece on this a couple of weeks ago on Lenny's podcast, basically saying that when you think about if delivery becomes very cheap and easy, what ends up mattering the most is taste, judgment. Do you understand the customer really well? Can you do those data deep dives? Do you have taste?, right? Um, can you, uh, execute things end to end? Uh, and that product management in a lot of ways might be more relevant than ever. Sanna, do you have any, anything to add?

Great response.

Yeah, I agree with you because you need to put yourself in the customer's shoes, right, when you design, uh, these tools. And that's a very human thing to do till agents are customers, right? Um, with, uh, with the trajectory you guys are on, I think you've put like your growth rate and your retention and in the podcast description. Tell me a bit more about how is it going right now. Is it satisfying your expectations and what's next for you guys?

Yeah, I think for us, we're really excited. You know, our DAU to MAU ratio is 40%, which is almost consumer application territory. Net retention has been great so far. Our week-on-week meeting volume growth rate is about 10% week-on-week. Um, and, uh, in terms of, uh, PMF signals, like, we basically run a, like, a superhuman-like survey, uh, for all of our customers, right? And then we have to ask a question of, like, how would you feel about, you know, the solution if it went away tomorrow? And, uh, you know, like, right now we're over 50% of respondents say that they would be super bummed if it went away tomorrow, right? Most people say they save about 5 hours a week. So, so all the signal is, is great. I think the big thing for us is, is, um, just augmenting go-to-market, right? Like, we feel like with our enterprise customers now, with the, the folks that we have like in the solution, uh, now it's more of like sort of how do we have a repeatable playbook, right, to get more of those opportunities in. But yeah, we're really confident about the product.

See, and then I miss anything from your last, uh, couple of roles, Mark, like, um, at the exit with ProductPlan, you took, um, Mindbody Engineering from 100 employees to 250, which is gigantic, $50 million revenue to $100 million with 6 acquisitions. Uh, do you feel that you're on target to achieve something, uh, greater than, than this without fundraising?

Um, I think that we can go really far, you know, uh, without fundraising. But I think, I think at some point, um, if we're realistic, we want to win this category. Um, that's kind of like the, the decision point for us, right, is, is category creation. We think category, category creation is possible. We think that a solution like this is inevitable, whether or not we build it or somebody else builds it. Um, we want to win, you know, so I think, I think we're going to have to, you know, capitalize, right, in order to win. Yeah, and I think the opportunity is much bigger than anything else we've been a part of.

I use Granola for my computer. It's pretty good. How are you guys different?

Sam, you wanna take it?

Yeah, so I love Granola. They're a great tool. You know, it allows you to essentially jot down notes and at the end of your meeting, it kind of, it flushes it out in a really nice format based off of what you care about. And what's really interesting about that is like the output of that is a very nice summary of what you've done, but you still have to essentially do the work after your meeting. And what I mean by that is, you know, say you're, you're a PM and you're in standup or you're running through a couple different engineering syncs or product updates, and some of the takeaways might be, hey, I need to add a couple Jira tickets or Linear tickets to the board, or maybe I need to set a follow-up I'm, you know, meeting with a certain cohort of people, or maybe I need to send a status update, you know, to my boss. Now imagine, you know, earmark having all of that context of your meetings, but also being able to essentially connect with all of your different tools that enables us to actually complete that work for you, or at least give you the first pass. So for example, when you come back to earmark after your meeting, It's like, hey, based off of your meeting, we think that these are the two linear tickets that should be added to your board based off of what you talked about. Does this look good to you? If you say yes, great, we'll take it and, you know, send it, send it that way. Likewise, if there's kind of like a follow-up email, hey, here's like a draft of something to send. Do you want it? Do you want to go? Awesome. So keep going. So what we are really trying to strive, or we're different, is is we don't just end at summaries, but actually end at that completed work that you would still have to do after your meeting, but almost free your mind from having to deal with that. And then you yourself can focus more on the strategic stuff and the more fun stuff that you care about in your role.

And what would be the future CEO's day in a life like? They would just spend their time talking to a computer and the computer would execute, send an army of agents to execute stuff on their behalf? Would that be the reality in 5, 10 years from now?

Yeah, I think that that's not too far from what it'll be, Charles. And I think that when we think about things that are uniquely human, right, or, you know, just how creative we are, and then voice is sort of like our own API, like as groups, right? We're tribal. Ideation comes from riffing, right? We all kind of feed off each other. That's where great ideas come from. Great ideas come from friction. Can you take, can you take that, that beautiful mess and discern it into things that are deliverable, you know, like straight away, right? And I think that's, I think that's, we're convinced that's the future of work.

Isn't it interesting also that we started writing, and that's like another specific region of the brain that has like also its own benefits when it comes to digesting and sharing information? Like, do you think that will mostly go 'cause I like writing. I, I think from a young age, not everyone loves writing, but more and more nowadays, especially when it's long messages, I use Whisperflow. I'm pretty sure you guys do as well, right? So any comments on that?

Yeah, I think, I think different modalities are gonna be required for different types of work and different types of thought. So to your point, like, you know, I think, I think, you know, concerted long-form writing and, and really fleshing out ideas. There'll be a place for that, right? Like if you think about if I was going to come up with a company strategy, right, or what is our narrative, you know, those things take time, they take work, they take iteration. Writing is the ideal place for those things. But, you know, the interesting thing for engineering and product deliverables is that oftentimes cycle time might be, you know, what used to be 2 weeks, right, in terms of an increment or a unit of value or work. And now that's getting down to days, hours. So voice is kind of like the ideal, you know, surface area for those types of deliverables, if that makes sense.

I would also add that— sorry, I didn't mean to cut you off, but I would also add that getting at a good starting point can be really helpful and you can take it the rest of the way there, like with human writing. So like what I mean by that is, you know, imagine you're writing a performance review and like getting that first step of like, okay, here, here are all like the things that this employee worked on. Here's all of the things that, you know, other employees said, basically like getting all of that evidence, like gathering, like pre-AI, that was, that was kind of a long, kind of difficult job to do. And arguably, like you should, you definitely should write that, you know, yourself, you should not outsource that to a to an LLM, but you can definitely use the AI to give you your first pass, and then you as a human can then—

what's the different shades of personality in writing versus talking about stuff? I feel writing, I'm like more discreet and less bombastic than when I talk. Um, I feel emotional, uh, there's emotional insight when you speak as well. And in writing, there's also like, if you dissect how someone writes, there's also insights to be had here. So what are the, the difference between the two in terms of, uh, yeah, qualitative information you can get from there is sort of my question.

Yeah, that's a good question. Um, right now in terms of what we are ingesting, we are pretty much only ingesting audio. And actually, we're not even actually looking at the audio, we're actually doing the transcripts and then actually getting that transcript. We haven't quite dabbled in actually getting any other types of like written text yet. Mostly that's kind of what Mark was alluding to, and like most conversations, you know, happen in meetings. But in terms of, I think, your question, I don't think we fully have explored that.

It's just an extra source of data for you guys, just like visual data, and just like Mark making the queues, uh, for, for UTS talks. And then, so, yes, I guess all like information you can digest, feed to an AI, and it can make sense out of it, right? Um, so, okay, um, well, Mark, you're on revenue, you're on hiring as well. I'm guessing you guys are hiring. Is that like part of, uh, next step as well, talent?

Yeah, so, so, uh, what's crazy is, uh, you know, being, being an AI-native shop for real. Um, I think for Sam and I, uh, we're gonna prolong the process of just us being the only, the only real, you know, bench. Um, and then if we think augmentation, it would probably be in the form of contractors for the foreseeable future until we get to like sort of a magic revenue number or when any discipline becomes so untenable that we have to bring on other folks. But I think, I think as a rule of thumb, automation of all workflows first until it breaks. Sam, is there anything you want to add to that?

There's so much that you could do now. Wow. That literally just changed in the last couple of days with Fable release. Fable. It's almost like every couple of weeks we just get more and more productivity gains.

Is that fun or confusing and scary for you? Oh, it's a lot of fun. Like the fact that they—

Okay. Yeah.

Yeah. That's what I would guess, right? For technologists, for most humans, it's scary though.

Yeah. Charles, I think for us, like, I think being on the other side, you know, doing product development in the traditional manner for so many years. And then now, like the fact that like we almost, you know, understand that like throughput is all limited to your imagination and what you can feed to agents. Is, is like the most wonderful problem ever, right? We're all rhythm players. Like, we just, yeah, we just love it.

Um, so you're aiming for a two, two-man billion-dollar company if it's possible?

Probably not.

Yes, let's, let's go!

Let's aim for that.

Um, so, okay, um, talent, yeah, that outsourcing to agents. Both of you, why are you guys like the best, uh, combo ever?

I just think that, like, you know, like, you know, like, there's a, um, you know, when you work with somebody long enough, you know, so to see this point earlier, right, like, we've gone through 5 tours together and we've seen scaled enterprise complexity, we've seen, you know, zero to one a couple times, and then now rolling our own. Um, I think, like, we've seen shit hit the fan, we've seen it all, you know. And, uh, and because of that, we know we don't like You know, and we know we don't like a lot, uh, and we know what we do like, right? So I think we're both very aligned and opinionated in terms of like, like what the company vision would be, like what is a great work environment, like the problems we want to solve, what we geek out on naturally, you know. So that's high alignment. Um, and then also, um, you know, it's like when you have a great band, right, that's been playing together for a long time, um, you know, at some point I don't even like tell Sanden like he has a solo the next bar, right? Look at him, he knows when to solo. There's a certain intuition that comes with that and a certain level of trust, and it's been great.

So, day in your life is like both of you on Slack, just exchanging messages, Mark, you more on the GTM side, Sanden on the coding side of things. Do you both share a dashboard in which you can see daily progress and that this thing is taking off?

Yeah, you pretty much nailed it. We essentially live on Slack all the time, and yes, we use PostHog, where we just have a ton of data showing kind of our live usage, showing the daily usage of our product. And what's really cool is that lately that's kind of been giving us insights into like where magic moments are in terms of like the customer, and then we can kind of make sure that we really bake that into the product. Like, how do we drive people into there? So, um, yeah, that's, that's, that's pretty much it. And, uh, we're—

so go ahead. Who's your most productive agent? Let's just keep it, uh, to that for the last question. Like, who's your, your best performing agents?

All right, in, in terms of, uh, just, just overall workflow Yeah, yeah.

So, so like, GTM for sure, it produces 20 meetings a day.

Yeah, it's, it's— Shannon is like the maestro, right, in terms of orchestration. He's got like, you know, multiple like Codex, uh, agents running concurrently on any given problem, right? Prototyping new problems, uh, roughing out old things. You know, oftentimes what will happen is we'll have, uh, um, something we really pride ourselves on is there'll be like a user-reported bug, right, or feature 'Cause we have all of our customers in Slack. And then oftentimes within hours, Sandman will have a solution. We actually had a conversation yesterday where one of our customer service Titan, they asked for audio notifications, like a little tell, when the Meteor miner comes up. We, last night, basically created a solution 6 variations of tones right now. We published it, what, like half an hour ago? So like, uh, yeah, so Sienna's orchestration is worth that.

There you go. Well, thanks for coming, guys. Uh, I was Charles Cormier and