Ignite AI: Dennis Mortensen on Startup Failure, AI Agents, and Why Boring SaaS Problems Win | Ep278

Dennis Mortensen: If I speak to any young entrepreneur who's a few years in and drowning a little bit, if I were to kind of give them one piece of advice is if you can just figure out a way to just not die, that's probably the best way you can somehow win. Because most people just die off naturally. As I couldn't get it to work doesn't mean that the idea, team, product, uh, or customers didn't exist. It might all have been in place, but somehow you didn't live long enough to see that flourish. So you can just arrange for you to not die, you might be winning. And there's plenty of times where not. I'm just not dying here. I'm just going to keep leaning in. The more I lean in, the more I become world class at this one thing.

Host: Hey everyone. Welcome back to the Ignite podcast. Today we have Dennis Mortensen on the mic. He is a Danish born New York City based serial founder who has sold four companies and torched one. He ran enterprise web analytics at Index Tools, sold to Yahoo. Predictive Media analytics at ah, Visual Revenue, sold to Outbrain, and then spent eight years building xai, which I totally remember from back in the day. Uh, an AI scheduling assistant whose email boss Amy and Andrew Ingram booked your meetings before most people knew what an AI agent was. He raised 44 million from Firstmark and other VCs, sold it to Pizzabo in 2021 and did this all, all years before LLMs made AI agents a dinner table phrase. He wrote a book, Data Driven Insights, published by Wiley, taught analytics at UBC and his first venture, Evonix. Am I saying that right? Evo Knox, a Grubhub for Europe attempt in 2002. That's super, super early. Went bust in what he calls a rather expensive mba. I love that. Uh, today he's building Launch Brightly, which automates the product screenshots in your help center. And he has very loud opinions about why that unglamorous problem is core AI infrastructure. He is a rare guest who can tell you from scar tissue exactly what building an AI agent costs and what an AI doesn't actually work yet. Thanks for coming on, Dennis.

Dennis Mortensen: Yeah, looking forward to it.

Host: Yeah. So I'd love to, I mean, I kind of gave your background, but I'd love to hear the, you know, between the lines, what's your origin story and backstory?

Dennis Mortensen: I think you did a very good version. There's the short version, uh, that I'll add to my title on LinkedIn. Then there's the, uh, four hour version which you can call my mom for, which is spectacular. And then I think there's uh, the kind version you just did. But the origin story is probably me deciding early on that I did not want to be an entrepreneur. All I wanted was to get my CS degree and go work for ib. So my whole family is the whole kind of nature versus nurture. We're entrepreneurs not in tech, but elsewhere. So my dad, uncles, cousins, what have you. And I see the whole thing up close, what it takes. And it doesn't matter whether you're in tech or not, it just takes effort when you try to bring something to life and that whole one story. And I thought, nah, you know what, I want to take my bicycle, go code IBM, had an office in my town, go code at IBM, get home at whatever, four, hack away on my laptop, be free on the weekends. And that was the plan. And during university I did game development. And we can talk about how programming anything have dramatically progressed. And we've certainly taken a step function of late for where perhaps the programming language of the future is English, or even poor Danish English, which I can kind of uh, throw at an LLM and get out reasonably good code. But did that and the company I kind of worked for went bankrupt. And I ended up for very little money acquiring the assets of that company because I was just so, to put it bluntly, and I hope this is not kind of PG13, but I was pissed that the year and a half that I've spent along with some other geeks working on this particular title would end up in nothing. So I bought the assets and ended up selling the finished game shortly thereafter. Made a few monies back then. I thought I made it big, you know. Now of course in hindsight that was not the case. But hey, I had a few extra monies in my pocket and I thought, you know what, let me go start. Uh, I saw some opportunity to do log file analytics, a firm around that theme. Spend all of it, keep any of it for a rainy day, all of it lose, then go work for IBM. And then of course you get started and you figure out, yes, we had a very good outcome. We did that for four and a half years. I'm one of those kids who got an exit on the right side of the dot com boom. So I sold in April 2000. So not that I've seen anything, the

Host: sky was already falling in in April. I mean I remember it was like

Dennis Mortensen: March if you look at the chart. But nobody thought that was just volatility. Uh, but you're right, I sold. NASA was still, I think short of 5,000 or something like, like that at that moment. So very good timing. But. And then you're at it and you figure out perhaps this is what I was meant to do all along. And done. Now I'm on my sixth venture. Did all of them back to back. Didn't, uh, do any.

Host: Hey, let's go sabatical.

Dennis Mortensen: No, just back to back. We leave Friday, we start on Monday.

Host: Yeah, shut down operations or sell off on Friday and like start the, start the new thing three days later. I love that. What's the connection? I see Hungary in the notes. There was an analytics company out of Hungary. My wife's Hungarian. That's why I asked.

Dennis Mortensen: So I spent four years in, plus four years in Budapest or on the Buddha side, by the way. The office was on the Pest side. So I've taken the dramatic hour and almost 30 minute commute. I used to live in the old.

Host: To go like one mile.

Dennis Mortensen: Yeah, it's not very long, but you only have whatever kind of four core bridges. So the whole thing is just a little bit of a nightmare. But yes, we had an opportunity there and built up some really good technology. So we did this before. Not before, but just on the verge of Hungary being a kind of true participant of the European Union. And it was really still a good outsourcing nation. So we didn't take any VC capital for that venture. Built it up from the ground up on cash flow and had a very good exit to Yahoo. And obviously on exit day. And I love my investors, it's. But it's certainly nice on exit day that you don't have to kind of split the pile, uh, into kind of multiple smaller kind of versions. You can just kind of, hey, we're these four guys. It's between the four of us. And the funny thing is, part of the deal terms on that deal was that I moved to New York. I pissed hard for the West Coast, Silicon Valley, all that good juice. But nine hours between Sunnyvale and Budapest is just, uh, too dramatic. And there's not really much of an overlap. So we're going to put you in New York, Dennis. We're going to keep the team in Budapest, but you have whatever, two, three hours with some overlap. And I thought, yeah, I'll do my tour in New York, then I'll move out West. Short of 20 years later, I'm still here.

Host: It seems to be the common path of an entrepreneur in, in Europe. Right. At some point you're kind of Looking

Dennis Mortensen: west, it's certainly hard not to look west just to the US As a whole. It's almost like, I don't know. So we have one sport back home, which is football. So you play for whatever local football team and that's nice, but if somehow

Host: with your feet, like actually with your feet. Football.

Dennis Mortensen: Yeah, Soccer. What do you call soccer over here? But I'm fighting an uphill battle of not wanting to call it soccer. And I'm losing, by the way.

Host: Yeah.

Dennis Mortensen: So. And I have to soccer all summer. It's not happening. So there you go. So soccer is the only spot, but it feels like you play for some club in whatever kind of local place you're at, but you have the opportunity to go play for some, you know, top six premier league club. That's what the US Feels like. As in if you really want to kind of just lean in, kind of play with other like tier one players, you almost have to go. I know we've come a long way certainly since I left, but still. And then when you hear on the east coast, there's still a little bit of, I hope, I hope this is not recorded, but there's still a little bit of kind of west coast envy some days. I'm trying hard to advocate for the east coast because I've been here for so long, but New York is kind

Host: of is the secondary hub. I mean it's. Yeah, like if you look at the, the data, I mean San Francisco is just far and above and beyond, uh, in tech. But you know, New York's a solid second, really came from 00:20, 25 years

Dennis Mortensen: ago, you know, and I, and again, I'm so biased, but I do think if you're doing anything in B2B SaaS or not, there's certainly some luxury of me being able to take a 10 minute subway ride and find the first 40 customers. As in the whole thing is just around me. That gives you a little bit of a commercial, I think, upper hand. I think compared to some of the work you have to do on the B2B end on the West Coast. If I were to kind of pull one attribute where I think we're doing quite well out here and in any vertical, by the way. Oh, you have to be finance, fashion, content. That's a customer across the street.

Host: Yeah. Well, let's talk about your grubhub for Europe with a twist. And you know, what you called the expensive mba. What's uh, a, what's a lesson, you know, from that failure that you still use today and how did that turn

Dennis Mortensen: out so timing was not. There's obviously many lessons and real scar tissue because I funded it myself. So when I talk about an expensive mba, well, I think, I think I could have funded a whole class of at Columbia for that mistake. So it's not just mine. I have a whole kind of class of whatever 2007 I paid for. But a few things. If I was do a couple of takeaways, I'm. That's a. There's a defect on me and I'm working on it. But I am somewhat allergic to the kind of dramatic pivot, as in if Brian and Dennis agree to go to battle, we can't halfway there decide, you know what, we're not going to invade, uh, Poland, we're going to turn left, we're going to do Sweden instead. Like, that's not what we agreed on. And the whole team I put in place was to go on a particular mission and us kind of doing the pivot to something so dramatically different to what we agreed on then this can't even be the right team anymore. You should just call it the day and say, you know what? We tried. It was now time. It was like good timing. We weren't the right team. There's such things perhaps even outside of, uh, our control. But we certainly didn't win. We don't need to pivot as. And I don't think the startup, like a traditional business, is something that must survive at all costs. I think it's an experiment for where you set out to run a particular experiment. If you prove that, well, that particular experiment concluded decisively that we shouldn't exist, then you just shouldn't exist. We'll get a paycheck somewhere else. So I'm a little kind of allergic. But that doesn't mean that you can't optimize on your idea. And perhaps the whole thing is this kind of, uh, vocab. But I was unwilling to change one thing which had the whole thing kind of collapse a hundred things. Right. Or I was just not good enough. Let's just start there. But what we did back then, which worked suddenly very well in ramping up revenue, is that we built not a marketplace. So today, if you go on GrubHub or any one of these marketplaces just eat back in Europe. It's truly a marketplace where Brian needs or craves Thai food tonight or burgers or a pizza, and you go buy something and you have a bad experience. Sure. They might even have some ratings attached to it. I don't want to Buy from that place, you buy from another place. If everything you get from the marketplace is poor, you stop using it. But not on one or, uh, two poor experiences. That's just that particular place. We actually don't really blame the marketplace. And it's just one for where. Nah, one style rating. I'm going to use the other place. Now. We did something different, which is that we created a. Essentially a marketplace. We're kind of playing out what, uh, the guy from Jet is doing with Wonder. So we would rebrand and kind of franchise it and sell a dedicated menu across a certain set of verticals, and we would own that. So you bought it from us. We obviously had people kind of producing it, uh, for us. But that means I own the responsibility of the quality of the product, and I am to. I am the only one to blame. Of course, I get a much better markup compared to. In a marketplace. You kind of live and die on fees and people fucking hate your fees. But here I could mark it up as a premium experience, and that worked, uh, very well. But I never figured out how to really craft the loyalty from the folks that produced the. The one thing I delivered what I thought incorrectly. Now, in hindsight, I still think it's a little true, was that anything which you order from grubhub is not supposed to be a culinary experience. That's supposed to be, I'm busy and it's 8pm let's just order something and get some food. And if it's not shit, I'm kind of happy. But you know what? It needs to at least cross the inflection point of not being shit. And these guys quickly figured out in this particular industry for where I got no loyalty to Dennis. So I can use, you know, anything subpar can go out on this channel, anything in my own name. I'll just dial up a little bit, you know, uh, on ingredients, delivery times and whatever other kind of attributes. So I died on that. But damn, we ramped up revenue fast because this was at the very beginning of when people started to kind of order things online, actually. So much so that we had kind of a merger opportunity with Just Eat, which is the biggest player in Europe, which I declined. We would had about a quarter or something like that. Finding my notebooks, I think 22.3% or something like that, which I declined because I thought they felt like amateurs a little bit. But they did. They fucking killed it. And we die like four months later. So I ramped up revenue dramatically and that worked very well. Uh, and I Thought like in many places that scale, scale solves many things. Here though, if, uh, it's forever 1 out of 20, which is not good if you don't solve that kind of underlying issue. Yeah, you can scale all you want, Dennis. You're going to have a problem just at 10x the scale. So we ended up dying on that and I called it the day. It was a good experience.

Host: Uh, I love that because as a founder, you kind of want to believe the long term vision and you want to stick with it. And you want to believe like, hey, no, we have a thing here that could be, you know, a billion or $10 billion company. And you really believe that, right? Otherwise you're not doing it. Or maybe you're not doing it. And so you get this acquisition offer, you're like, hey, we're a competitor, we'll buy you, we'll give you 22% of the company. And you're like, no, I think I'm going to go it alone. Like, if you could go back. I mean, obviously hindsight's 20 20, but like for founders out there listening, when do you know it's the right time to take the acquisition offer?

Dennis Mortensen: You know, so there's certainly two schools of thought. So my mistake, by the way, was not not taking the offer. My mistake was not seeing more clearly that I'm actually in a space where it's probably going to be dominated by marketplaces. We'll see now if Wonder becomes a thing, then I was just, uh, 20 years early, if you've seen what Wonder is trying to do. So they're trying to kind of. They're also owning the restaurants themselves or kind of the cloud kitchens that they,

Host: you guys were vertically integrated and making your own food and delivering it. So you, you guys had better margins. But it turns out people wanted more variety. Right?

Dennis Mortensen: So the, the, the mistake was not not taking that. My mistake, as I see it, was one for where. Well, the market, no pun intended, is trying to tell you something, Dennis. You should see it clearly. And your vision was not necessarily one where you should have been married to the rebranding. It should have been one for where you've been married to the idea of delivering food online in a setting that might end up being a marketplace with multiple brands that are not your own. Hell, I could even had just mixed it. Just let your own branding compete, Dennis, if it's so much better, well, it'll win out in that marketplace that you just happen to own and operate. Shit, you can even do a Google and kind of have your own kind of products, you know, arrive first in the search and so on and so forth. Right. There'll be all sorts of things where. If you can't even win, then Dennis, then the marketplace is just what people want. So there's just plenty of things. Now if I could go back and play it again for where? Don't be. Don't do the pivot. Don't do the aggressive pivot. As in, I don't think that whole Twitter story is sexy. As in, I'm not dazzled by audio becoming Twitter. I think that whole thing is just luck. And hey, it's a great story. I don't think it's useful for anything.

Host: Maybe you could really quickly tell that pivot story for folks that aren't familiar with that, because that was 20 years ago.

Dennis Mortensen: Yes. Yeah, I should pick stories that doesn't tell everybody how old I am. But the whole idea of them having done very well on creating Blogger, this first is real serviced hosted blogging service where you could just sign up, get a blog up and running, and be off to the races. And did so well. They ended up getting acquired by Google. Spent some time on Google, obviously had a lot of glitter attached, uh, to them. Evan then walked out, left, created this new company, Odiyo, which was supposed to be a podcast type platform with a lot of good ideas that, by the way, now are true back then. Weren't true just yet. But, uh, he obviously had a picture of the future that eventually got realized, but it didn't. Didn't really work, and then ended up hacking away on this internal idea that became a Twitter, which they then spun out. There's a lot of, kind of debacle attached to kind of whether the constituents on both sides got what they deserve. But that's not the point. The point is that going from a podcasting type service to Twitter, there's no overlap in that Venn diagram. They're not even kind of on the same piece of paper. And that, to me, is just not something which I'm fond of. Doesn't mean that I couldn't fall in love with the Twitter idea. But then I take me and a set of friends, make sure that we close this thing down, we walk away, and we build a new team on a new mission. But not one for where, yeah, we got a little bit of capital left. We need to kind of figure out how to kind of eat tonight. Let's build some new products. Doesn't ring true to me. So that was my, my learning, which is that that was not A audio to Twitter pivot. We should just have been one for where we're still doing all the awesome things about building infrastructure for how to get, you know, you being hungry to food being on your table. All that infrastructure is still intact. It's just that the little logo in the upper right hand corner in what I sell is not my brand. It's somebody else's brand. It's not even a pivot.

Host: Well, I think, I think. Who does that now? Like, uh, I think I was at aws and they do that pretty well. Right. They have first party services that compete with their third. Third party Marketplace services. Right.

Dennis Mortensen: That's a very good analogy. That's so that's exactly. I think what they also solve.

Host: And you can now tell me Amazon Basics is another. Another thing. Right? It's like you could buy the Energizer batteries or you could buy the Amazon Basics batteries, but either way, we're making money either way, you know?

Dennis Mortensen: Yeah, so. So I, I do, I do. Some days, you know, you walk the streets or look out the window, you have a DAI code. I play that game again and again. Like, where was it for where this would have been obvious. So these guys obviously wanted to merge with us. So it's not like the idea was so foreign that I couldn't imagine it. No, it was, it was painted for me. So it's not like, oh, I've seen it. So I must have had a moment of deliberation where I then decided, like, what was it? What was it that had you say, Dennis? No, I'm not doing that under any circumstances. Which we didn't.

Host: Yeah, I mean, it's sort of like when you're a founder and you have a certain way of going to market and this certain business model, you're sold on the vision, right. And you're trying to sell everybody in your organization. You're selling your investors, your customers. And it's almost like you have this reality distortion field and almost tunnel vision where you have to keep going on the mission. Because if I admit that some part of the mission's wrong, then I was wrong. And there's sort of like this cognitive bias that kind of creeps in where you're just like this like sunk cost fallacy, maybe, or there's a thinking bias.

Dennis Mortensen: Yeah, it's probably some sort of amalgamation of multiple biases. And perhaps this is a distinct founder type bias that somebody at some business school will go do some research on, because I think it's. It's certainly its own distinct flavor because it's not black and white because there's also, this is a good little story of me not doing the right thing. But there's uh, certainly also plenty of times where not surrendering to something and sticking to your guns and keep leaning in is exactly what you're supposed to do. As in, if, if I speak to any young entrepreneur who's a few years in and drowning a little bit, if I were to kind of give them one piece of advice is if you can just figure out a way to just not die, that's probably the best way to, you can somehow win. Because most people just die off naturally couldn't get it to work. Doesn't mean that the idea, team, product or customers didn't exist. It might all have been in place, but somehow you didn't live long enough to see that flourish. So if you can just arrange, uh, for you to not die, you might be winning. And there's plenty of times where not. I'm just not dying here. I'm just going to keep leaning in. And the more I lean in, the more I become world class at this one thing for where, yeah, I could have done seven things, been kind of half assed at uh, those seven things most likely because it's so hard to be world class at one thing. But if I just do one thing and keep doing it, it'll be obvious to people that if this is of value, well, Dennis fucking nailed it. Like he is the expert. The market might end up being smaller, uh, larger, uh, or whatever. But the whole thing is that, yeah, that thing, he owns that. So, so it's not black and white.

Host: And then so, um, so you end up shutting it down. What's, what's that like four months later? Um, maybe you can talk about the wind down process and, and how you kind of transition into index tools.

Dennis Mortensen: So I, so what I did do is so I called up these guys as in if you have any pride and you're not willing to surrender your pride, you're probably, uh, probably not going to do any ventures because you'll have to surrender that pride often, especially early on, on many fronts, probably all fronts, on raising capital, bringing in new candidates, speaking to the press, speaking to your employees. Like there's just so many things for where it's not going to be as pristine as you had hoped for and it's going to be a little bit embarrassing, but if you can't swallow that, you're going to die. So I just called up these people and said, hey, you won. We could all have a Laugh about it, but today I'm calling it quits. So you're going to buy a few assets from me. So I get a, uh. This was not an exit. Just. So for the record, you're going to buy these assets for a few monies. You can take on these employees as well. I'm going to keep my own laptop. I don't have the energy even to reinstall a new laptop. So I'm just going to take the laptop as well and then I'll be off to my next thing. So that's what I did to close it down. They took on the engineers and some of the tech and what have you and even did a play on then incorporating this into their marketplace. The whole kind of thing we just talked about. So that was the kind of wind down in, in that regard. And yes, it's not even like I turned up in my suit or anything. I think I did that call in my underwear in my bed looking at the ceiling. So it's not like. Do you know when you just. Yeah, this, this is it backing out of bed. Yeah, it looks like you laughed and it's like visually as well. Yes it did.

Host: Mhm. And tell us about index tools. So that uh, how did, how did. What's the story there?

Dennis Mortensen: So this was the moment where it was not immediately clear, but sometimes you have an inkling, right. You could almost uh, taste it. And I don't have it fully formulated just yet, but it certainly seemed almost inevitable that anybody doing any commercial endeavors on the Internet would have to kind of track those endeavors and see if they deliver the results they had hoped for, whether that is a, you know, a marketing spend or a build of a feature and so on and so forth. And as that happened, the prior setup to do this, which was doing analysis on your log files from your web service, which we had done prior, so we had like an intimate relationship with what it would take to, to extract some insights from web traffic. We just done it with a like three week latency, not a two second latency. Right. I said, you send me your data on CDs, I extract that onto a set of servers, do the analysis and I do a handover kind of McKinsey style. But it seemed like we now progressed to the point where we might be able to do this in near real time. Things that, you know, Brian and Dennis can spin up a Mongo instance tonight on some server and do pretty much real time analytics, not so much back then. So we ended up seeing that as an opportunity where perhaps there's a moment where not only do you not need it, you can get the marketeer to add a pixel to your website. That pixel becomes the tracking mechanism. And again this is all very standard in the year 2026, but not so much back then there was a whole kind of beginning industry and we built that. And without going all geek on you, but damn, we built some, some cool technology. So this is again I'm just showing how old I am. But this is back before kind of SSD disks. So these will be kind of magnetics spinning disks for where a uh, lot of the time you would spend would just be loading up the data itself and people might even remember that you would actually buy new disks.

Host: I remember going to Boeing back in Seattle in college. So this was, must have been 2002 and they took us to the data center at Boeing and it literally had a robotic arm going around the data center and it would grab 20 gigabit gigabyte disks literally uh, like off the shelf and plug them in so you could read the data. And these are 20 gigs.

Dennis Mortensen: So it was just. But so we ended up engineering our own file system because many times when you have to read from disk you end up spending a lot of time just spinning so I can get the next kind of chunk. So we did chunks. That was exactly one spin on the disk. That's not useful for Windows or any other operating system but it was useful for us if we want to load data rapidly. So we built our own file system, then we built our own database and then we built our own query language. So we built a column based database. So almost all databases obviously row based as uh, in select Brian from customers and then you'll get some sort of row from some sort of table. But in analytics of course you're aggregating things so you're mostly going to try and kind of query down columns. So we built a column based database, uh, and then a query language to kind of fit that and ended up with about 7 second latency and kind of really nailed it in that regard because that was unheard of at that point. Again, you and me could do it tonight alone for uh, a hundred bucks.

Host: Right, right. But yeah, well there's all kinds of key value pair databases and Vertica, Vertica comes to mind from 10 years ago and stuff like that.

Dennis Mortensen: Yeah, but very sexy. Built that up and then built the kind of front end analytics application on top of it to do clickstream analytics. Uh, and ended up getting a host of customers. Did very well, Became cashflow positive and had multiple kind of offers along the way and were actually in the midst of a signed letter of intent type due diligence when Yahoo, uh, came in. Here's a side note to that. So there's two schools of thoughts here. Almost kind of coming back to one of your earlier questions for where, when do you know it's time? That's what you ask. And there's the kind of wise school of thought which is that investors are uh, borderline evil. You should never speak to them. Corp dev are the worst of the worst. If you speak to anybody with an M M and a title in their setting, that's like just run away screaming. I'm not so sure I'm a believer of that line of thinking. I, I'm more a believer of the well, you should probably speak to everybody all the time. Not to the degree for where you end up not doing work you should, you're supposed to do. But that's rarely the case. So I do the opposite. So I speak to probably two or three every week. Uh, keep them abreast of what we're working on, where we at, whether I delivered on some of the promise, I've told them in kind of prior conversations. So if I one day need a friend, they are friends in the sense that we've chatted many times before and

Host: if I, uh, look at you have a champion there.

Dennis Mortensen: Yep. Like all, by the way, the four X's we've done have all been outputs of this process. That doesn't mean that that is the process. That is just the process for me. So on Yahoo, I've done a not a pitch but demo slash mini pitch for RBC. They didn't really invest anything less than $50 million. We were possibly looking for $5 million. We ended up not doing anything, just doing cash flow, but did it anyway. Not as in Dennis, that's going to be a complete waste of 30 minutes. Nah, I'm going to plant a seed then. Of course nothing happens here. First of all, we don't have wouldn't know how to spend $50 million. Perhaps we could have figured that out but didn't do anything. Eight, nine months later, it happens to be that RBC is the investor and bankers for Yahoo who were in the midst of looking at another company where ah, uh, there was a little bit of friction and they said, you know what, I uh, did speak to Dennis, I think we should go have a look at least, or at least compare. And that whole thing came about because I did a pitch. I was Supposed to do, supposed not to do. Right. Because of some kind of line of thinking that I kind of disagree with and probably did it in 10 days from initial high and low to having the whole thing signed.

Host: So the takeaway there is like if you're a founder, take the M and A call, right. Don't be ready to sell. But just these are relationship building exercises that could one day lead to an acquisition down the road.

Dennis Mortensen: And they're not. And again I call often bullshit. On the whole they are just a time sink. First, how many inbound calls do you get for this to be a time sink? Sure, if you get 30 a week, something else is going on. Sure. But then I think you have all sorts of other problems like good on you man. But you're probably not going to get so many that you can even call this a time sink. If you do two or three of these calls a week. I think you're just kind of preparing yourself for a future where you have much more optionality. You might not take any of the options presented to you, but at least you have optionality.

Host: I love this lesson here because I've been thinking about this a lot in my business. I made a few decisions recently that decreased my optionality and I won't get into what that is but I, I, I, I reminded of myself, I reminded myself that you have to make decisions that increase your future options you want. Yeah, this is a great lesson and

Dennis Mortensen: I'm just because I'm trying to sure people will come up with a few pointers on what the possible penalty can be but those almost feel contrived to me. As in sure you can say that but that I'm not buying. If I look at X AI when we got acquired that came about because there was a smaller startup who thought we might wanted to acquire them. I'm not a big fan of small uh, medium sized startups buying other small medium sized startups. It's kind of like multiplying 0.8 twice over. That's just a smaller number now so don't so we'll eat run to end but I took the call anyway. Not going to disguise anything said this is my opinion. How about you going to have a chat about what you have. Worst case you'll know I know what you have. Should I meet somebody I'll be happy to kind of share that information. Now nothing came of that. Nothing was supposed to came of that. However, half a year later uh, they all, they of course did the rounds, didn't really have the right technology were A little too early. They spoke to somebody else where, you know what? Dennis was a nice guy. I think they have the technology you are looking for. You should give him a ring. Well, that was the lead, as in, I planted my own ambassador who then went out and said, yeah, I know what Dennis is working on. If that's what you're looking for, give him a ring. And he did the introduction and everything. And that ended up becoming the, the deal that we did in the end.

Host: That's really cool. And then moved on to visual revenue. Five. Five years. Yeah. Uh, did you have golden handcuffs at Yahoo? You had to kind of stay there for a year or two or so.

Dennis Mortensen: We can talk about that. So we been very fortunate that we've not had any lockup, earn out or, uh, golden handcuffs in whatever the incarnation to any of our exits. So we've been able to walk out the door the next day without any financial penalty. Of course, if you do that, you'll forever be locked up in the future. So you can only do that once. So that's the last party you attend. But if you do not do that, as in when we got acquired by Yahoo for where there was no lockup, we stayed put for 17 months to make sure that the very thing you acquired, which you paid $10 for, feels like $12. The day we leave Bisabo stay for 14 months, making sure again, you gave us a few monies and I. And again, I think you can be very kind of upfront about this. For where, hey, you can look at me. There's no version for where I now wake up and aspire to some corporate career. And there's nothing wrong with a corporate career. It's obviously something for me. So you need to tell me how do I best make you look awesome in having us acquired? Uh, so what does that board meeting look like in 8 months or 14 months or what are the certain inflection points where once I deliver on that you look like a hero, then I'll just work my ass off to make sure you can walk in. And again, as they do that, I just plant seeds because these will be investors or board members where, yeah, we did acquire that piece of technology from that Danish dude. They were stand up people, hung around, made sure the whole thing got well integrated. So no logouts, uh, but stay put, which I think is the right thing to do.

Host: And if, you know, uh, I'm wondering if you ran into some of my Rocket Fuel co founders there. Richard Frankel and George John and Abhinav Gupta. Uh, they Were the uh, the Rocket Fuel guys. I think they left around 2008, 2009 to start Rocket Fuel.

Dennis Mortensen: Yet we must have overlapped then.

Host: So I joined Rocketfield much later, right before the ipo. But yeah, they were at Yahoo when you, when Yahoo acquired, kind of in the same area in ad tech, you know.

Dennis Mortensen: Yeah. So, so we obviously had all sorts of integration. So we did a lot of ad tech type analytics. So you spend $10 on Yahoo or Overture or on um, Google and AdWords. How can we attribute that spend to some success on your site? So we did a lot of overlap with all of those folks. And the funny thing is we did that deal directly with Jerry when he was back. Just for a short period of time, if you remember.

Host: Yeah. And then so you're there 17 months, um, and you're starting to get the entrepreneur itch again and tell us the kind of the origin story around visual revenue.

Dennis Mortensen: So we all have kind of processes in some sense for how we try to latch onto some new idea. I'm not immediately a fan of uh, Dennis and Brian, Some Diet Cokes, a pizza, my whiteboard. And we started kind of just fooling around with cool things. There's, hey, it'll be fun. There might even be kind of half of a good idea in there somewhere because we're not stupid. But they're not. Or they're really real pain points. They're just cool shit we would like to work on.

Host: Kind of like starting with the uh, technology and ideas first.

Dennis Mortensen: That's what happens when you do the whole whiteboard session. I like to do the opposite, which is that it's not kind of an immediate itch. It's not something which I can turn on. So I try to just live my life going about doing my things. But if something annoys me, I have a little list of hate on my phone, which is what I just add to all the time you're standing in line. A Citibank. There's some particular thing you need to do you couldn't do online and you're thinking this is shite. Let me add that to my little list. I order something from the Internet, it arrives. Something is not working as I had hoped for. I hate that I add it to my list. Just little one liners of things I hate. Because I find it very unlikely that I'm the only person in the universe that dislikes this particular experience. Now I do it for two reasons. The one reason is I feel truly almost unfaithful. Um, if I work on other things while I do my venture as in I am very unimpressed. All my friends does it. So this is not. They, they know I'm unimpressed, so they don't even talk about it to me. But I'm unimpressed by, yeah, I've done these kind of angel deals or yeah, I'm sitting on this board or uh, yeah, I'm an advisor to these two companies. How about you do your venture and saying, I thought that's what you're supposed to be focused on. So I don't do anything. I don't do any advisory, no board seat. I never cut a single angel check. The only angel check I'll ever cut is if I hope my daughters will do some venture. Sure, I'll uh, cut that check. But I do nothing. And that feels, and that's just the kind of the easy things to cut. Then there's the parts below where as entrepreneurs, right, you and me, that's kind of cool. Then all of a sudden I spend the whole weekend kind of hacking something together. And it is fun. It's just not the mission you're on. So I immediately cut that out and it feels like I. They're all probably just little shitty things. But I saved that little thing, put it on my list. I never look at the list now. Then I get to the end and then I'm allowed to look at my list. Now I'll have 5 years of one angry boy having added one line at a time. This time I think I had 17 pages of things I disliked in the universe. And I just kind of starts going to go through the list like, yeah, I think, Dennis, this is not even a pain point. This is just you being angry. Delete this. This is a pain point. And somebody did solve it. Good on them. Ooh, this arrived multiple times. Somebody's, this is not solved yet. Why is it not solved then? There's a whole process for how I do research on those kind of one liners, uh, and try to kind of qualify them. I'll give you one other point which is too many people apply too much energy and trying to prove themselves right where I like the idea of uh, doing the opposite. So I tend to invite, depending on the particular thing which I've researched, a handful of people could be some scientists, could be some engineers, could be some commercial folks. And I, uh, bring them in for a multi day agenda. Could be three, four, five days visual revenue. We spent five days, by the way, fly them to New York. Super kind of cheap compared to me spending four years of my life figuring out I shouldn't have done it. And their only purpose is to kill the idea, come over here and shit on it. Because if I can't spend three months and really prep and you can come in, surely as an expert, it might be in the language space. And I bring in somebody from academia who's done LP for the last kind of two decades. If you can come in just willy nilly and kill the thing, I'm glad I didn't do it. Uh, but then it's on me to defend it. It's almost like a mini roast. But if I can survive that host of days, well, then I can go to my next step. So that's how, uh, visual revenue arrived, uh, as well through that whole process.

Host: That's really fascinating. And there's just so many good nuggets for founders out there thinking through their ideas and what they should work on next. And one of the threads I want to tug on that you said is you don't do any angel investing or advisory. Tell me more about that.

Dennis Mortensen: It is so unlikely that any one of us can bring any new thing to life, that the default output is supposed to be that we die a sudden, instant death. That's the expectation. So given the very, very low probability, it seems odd to me that I will take that probability and squeeze it even more. Because you can't tell me that. Sure, you cut an angel check. Unless outright philanthropy, where I'm just cutting check to be kind and I don't look at it. Sure, I'll donate here and there, but here you probably need to have a little bit of a funnel. Like you need to at least do two to pick one. If you pick one, that's not a portfolio, that's still a gift. Then you probably need 10 to do 10. Then I need a top that's a hundred. Shit, that's.

Host: Once you get started. Once you get started, you have to keep going.

Dennis Mortensen: Yeah.

Host: Because then you have to build a portfolio. And this is What I tell LPs, that, you know, because Team Ignite's predicated on thousands of people helping our startups. Right. And I'll hear this when, um, I pitch my fund sometimes from LPs, they're like, I'll just do angel investing. I'm like, great. Are you looking at 10,000 deals a year? Are you picking the top? Like, you know, 50 a year, 100 a year that we're picking. Are you, are you willing to dedicate, you know, 50, 60, 70 hours of your week, every week to meeting new founders and picking the best ones. Because if you're not like, you probably shouldn't even, like, do it, then it's

Dennis Mortensen: just buying lottery tickets. And hey, if that's what you want to do, why just buy into the lottery then? So.

Host: Or just invest in a fund. Right.

Dennis Mortensen: If you think. I think this particular vertical is something for where in my distribution of the few monies I've made over the years. I want this allocation and I like this vertical.

Host: Yeah.

Dennis Mortensen: Then you find, you know, be an lp, you find a couple of good funds and go do that.

Host: And then tell me more about the advisory thing. Is it kind of the same logic?

Dennis Mortensen: Same logic, which is I do many things that are. This is not necessarily good advice. Uh, but I do many things and then I just, I, I stick to it because I, I really believe in it. So I don't like the idea. Like I'll, it's not like I haven't walked up and done a talk at nyu, uh, Stern or Columbia or for, you know, techstars, uh, accelerator.

Host: Yeah, yeah, yeah.

Dennis Mortensen: I do that all the time. One off, ad hoc. I don't like the idea that my co founders, my employees, my investors, one day they sit somewhere, they see my face on some fucking slide that is not attached to a logo of mine. Like, that's embarrassing. Like, I want. When you see my face, it's the, it's the company I work on and that's it. You don't see it anywhere else. It doesn't mean I, uh, don't do the walks, the talks and what have you, but I'm not attached to anything. So that if I lose, I'll take full responsibility. If I win, I'll take a little bit of credit. Like, you know what? This is, this is all I did, nothing else. There's no distractions. It's not like, yeah, you have all these kind of, kind of things in the fire. No, I don't. It's the only thing I work on.

Host: Yeah, I love that. And so let's wrap up the story of visual revenue because I want to get to Xai.

Dennis Mortensen: Uh, yeah, so we got exposed certainly to Yahoo at that point, who ran content in a way more scientific way to that of most of the other publishers. And they have massive volume still to this day. But suddenly, uh, back then, massive volume

Host: still to this day. Uh, yeah, they, they do a good job. I, I worked in content services at Microsoft, was my last job at Microsoft as I was kind of slowly sliding into venture capital, getting sucked into vc. And I worked on, uh, the rebuild of MSN Money. So Yahoo Finance, I'm very familiar with the Yahoo Finance as, as a content engine and a competitor. They're very good, very good at what

Dennis Mortensen: they do and they now done it for decades. So they certainly built some muscle around that. And as we were kind of fooling around with the content ourselves, uh, we certainly measured content. It kind of felt odd to me. All the customers that we had in the content space, not the E tailers and what have you, they were super sophisticated because, you know, money is at play here. But the content folks, it felt like most of their decision making and I felt because we had a hard time selling into them as well. Much, much easier to sell into Etail. I was like, it was very easy to sell into them but it felt like most of their decision making was kind of based on Diet Coke and gut feel. So that was kind of one of those little notes I took like, we got an awesome product. Fair enough. You're buying my competitors, but you're not even buying my competitors. You're not buying anything. Like you're fucking winging it. Like, sure, your product might not be a thing that arrives in a package in the post, but it's still a product that you're selling that needs to be consumed and you need to be way more scientific. Then we ended up being acquired by Yahoo. Then I saw some of the kind of science they applied to it and then I took my little kind of list of hate for why, why is this so shitty? So yeah, there's something here. Then we ended up essentially building a, if you think about it. So if you search anything on Google, there's a marketplace right on, on AdWords or Google Ads as they call it today. And, and they're obviously super sophisticated about how they run that. Uh, they have both the inventory, uh, and the marketplace. But it's kind of the same. If you go to CNN.com there's about 150, uh, positions on CNN.com that's the biggest hero. There's some smaller ones that are just links.

Host: You got the sidebar, you got the interstitial, you got the footer, you got all these different ad places, placements.

Dennis Mortensen: But that's really. If you.

Host: Which has gotten ridiculous by the way. On mobile. Have you like tried to read an article on mobile? Like I can't, I can't even read it.

Dennis Mortensen: That just didn't work.

Host: And this is, I worked in ad tech. It's, it's, it's insane. It's like the whole page is ads and I'm like, I can't even read the article.

Dennis Mortensen: Google's right, gonna solve some of that with amp, but you know, we can park that. So what we saw was certainly that you pretty much run that as you run either the, you know, station on the airways or the printed paper through a set of editors. Meaning that they get to decide what is on the agenda. And there's all sorts of reasons for why you should continue to do so, but you probably need a little bit more science than you just kind of imagining that this is what needs to be on the agenda. So we turned that whole set of uh, kind of front pages, that's the kind of the finance section, the sports section, the homepage, into essentially a marketplace where you own and operate it, but you also provide all the inventory, which is that your inventory are uh, the stories. So in essence, if I cut it down short, I figure out the value of each spot on their homepage for each kind of minute of the day. And I could tell you, well, cnn, it might be they had all these kind of self fulfilling prophecies. Yeah. But the one in the hero is now the most viewed story we have. No shit, Sherlock, because you put it in the hero. What is more interesting is, is there another story that is a little bit more hidden that is outperforming its exposure that might do twice as well if you promote it to the hero spot, if it's within the editorial guidelines and that we then turned into this recommendation engine where we helped them essentially become a machine editor. But as decision support they could take our recommendations, which they then did hundreds of times a day and built up some very good technology. Probably had 80% of the top 150 publications. As in that's not a newsroom I've not been in. And uh, that's not an editor in chief I, uh, probably haven't chatted to.

Host: Yeah, so it's like a publisher side analytics and optimization tool that allows the publishers to surface the best stories in the best ways.

Dennis Mortensen: Kind of like a Bloomberg terminal, but for content. For where uh, we're not automating anything, but I'm giving you insight what's new to you that you should take action on.

Host: And then eventually sold that to Outbrain

Dennis Mortensen: who did work on the article pages. And the way we got acquired is that I did a roundtable for, for some accelerator of some sort. Again, one of those half hours where yeah, it could be a throwaway, but a room of possible, uh, are there. Let me go do a talk. Albright happened to be in the room, the coo, David. I obviously did my song and dance. Uh, as a founder, you got that song and dance down hard. And he took note and kind of felt. Yeah. So Dennis owns and operates the homepage and section funds. We own and operate the article pages. Something here ourselves. Or at least have a chat with Dennis. And we had a chat, ended up doing something.

Host: Yeah. Uh, I love this list of pain. Right. You said founders should lean into the pain because the pain, if true and honest, never changes. The technology changes, but the pain doesn't change. Right. We have these things that need to get done, these jobs to be done. I have this pain. What was it about the scheduling thing where you're like, this needs to get solved. And I'm going to do it because that's a hard problem. That was really, uh, a hard problem back in the early to mid 2010s.

Dennis Mortensen: And there's a key word there in what you said. So this is not entirely true, but it's a simplification. So as a founder, uh, you tend to attack one of two things, rarely both things at the same time. It's either a market challenge or a science challenge. So the self driving car is not a market challenge. It's not like we are reading articles in the Wall Street Journal suggesting that. Well, I'm not so sure. Do we need him? Do we not need him? Um, we all just agreed. If those things can drive themselves, I'm fucking in.

Host: Yeah, but we can't. A month. 100% for both my Teslas. Yeah.

Dennis Mortensen: So, but now we, we're on the verge if not having crossed into the territory, as you suggest. Yeah, we figured out that was a science challenge. Airbnb, that's not a science challenge. You and me can build that site tonight. It'll look a little shittier, but we can certainly get the site up and running. That's a market challenge. Do I want some, um, murderer, uh, sleeping in my room next door, or can I somehow believe that the ratings have him not being a murderer or whatever the kind of challenges is that they figured out, but don't attack both at the same time for really good advice.

Host: This is really interesting. So as a founder or an investor,

Dennis Mortensen: if you're looking at an idea, you need to know, what am I even attacking? I get skeptical when my entire friends can't tell me. Which side are you on?

Host: Yeah. Is it, is it a go to market challenge or, uh, some sort of a market challenge or a technology challenge? I love that framing.

Dennis Mortensen: So for X AI, we're In a very similar setting for where, first of all, it was on my list of hate many times. I've always been, uh, I say too frugal, but the right wording is probably too stupid to hire an ea. So through all our ventures, whatever the size, whatever the round, I'd rather hire another quarter of an engineer than have an. Have an ea. I'm not even sure that's true, but, hey, that's me.

Host: Do you still feel the same way, like ten years later or.

Dennis Mortensen: But I'm not advocating it strongly, but I, uh, still don't have it in me where there's just too many intricacies for how I want to run my time and I become old or confident enough to not let time just disappear. Right? So I can control my time.

Host: I find when I hire people at Team Ignite, if they're, they're much better if they're very focused and EA is too general, like, you have to give them like, like, like you do this like you do, you read the updates from the founders, you figure out what they need and you help them get what they need. Right? It's like a very specific routine thing. But if you give somebody like an ea, like a general, like, hey, everything I'm looking at, you should look at too.

Dennis Mortensen: Be my help, be my helping hand. Yeah, Just didn't sit well with me. Doesn't mean that I haven't hired people to do all sorts of things that I initially did that we always do as founders, where I've kind of, um, outgrown that. It's not even my expertise anymore. We craft a position or the position is well known and we get somebody to kind of do that. Happens all the time. But the generic ea, or the personal assistant, if you want another word for it, didn't sit well with me. But one of the things that I disliked it was not necessarily the process itself, which is a little cumbersome, but it comes in small increments of two, three minutes of suggesting Brian, when we can meet, uh, was one for where it felt like he. Things got lost. So I want to speak to Brian. I did my part. I suggested some time Brian got busy or, uh, things happened and then it disappeared. Right. I'm an inbox serial guy, so I go to bed and there's zero emails. So once I sent Brian my part, I've done my part archive, so this is not lingering. It's now on, um, Brian to make sure that this thing will get scheduled now. Then things happen. So that I disliked, like, you know what I Did want to speak to Brian. What did not happen. Somehow, you know what, it disappeared. So it was certainly one of, uh, wanting that not to happen. And I like the idea of being able to kind of having it be done for me with very kind of anal specifications either behind the scenes or in the dialogue itself. And whenever I spoke to myself or spoke to folks, their fantasy was always the same. Hey, you know what? My hopes and dreams is that I can climb, uh, the corporate ladder to the extent that I become SVP of Flimflam at Time Inc. And at that moment I will get Tim my personal assistant, and then Tim will do this job. And it was the same fantasy for everybody. Somehow quickly became clear that we, uh, all have the same pain. Everybody disagrees that the meeting part of getting its schedule is not the fun part. Speaking to Brian, that's the fun part. If I can somehow escape that, that would be lovely. But the only imagination I have is Tim. As in I just need to kind of keep climbing. Then someday I'll get Tim.

Host: Someday I'll be able to get that EA that can do all the scheduling for me. Right.

Dennis Mortensen: But I need to be at VP level or whatever type of organization I'm in. Right. So just keep climbing. Because I'm certainly not as a kind of senior manager, uh, in this company getting an ea. So just add some years. So they all wanted the same. So it felt like this might be a worthy endeavor. And I can give you a few of the kind of tests we did, which was fun a little bit. So me and my co founder, uh, Alex, uh, initially just to see can you even schedule meetings with little context. So we became each other's when we did the integration over at Outbrain and we started kind of fiddling with this idea. We then we didn't have any EAs over there. Then we became each other's assistants. So I would schedule all of Alex's meetings. We just set up a shared email and then he would schedule all my email, uh, all my meetings, which worked like I hated doing his meetings. I loved having him do my meetings. Okay, that worked. Then as that worked. But we might know each other a little too much. You know what, let's hire an ea. Then we hired an ea, but not for us. Then I found a dozen or ah, so friends at director level. I know you don't have an assistant and don't bullshit me, but I've hired one for you on my D. She'll get this uh, done for you. So you can just email anything in regards to scheduling meetings. We ran with her for two or three months. Just going to see does things start to cluster? Because this could be one of these kind of whole world problems for where you solve one thing that showcased that you have another two problems that showcase you another four problems. They have another eight problems. And all of a sudden what you have is the need of a human. So does this at some point have a drooping.

Host: Can you do it with software? Basically you start with the human and let's, let's, let's, let's wyszy wig it right. Let's wizard of ait right.

Dennis Mortensen: And yeah, didn't mind and did that uh, and that showed you know what she even without us asking started to make macros. You know what? A lot of this is the same. Getting Dennis and Brian together, a lot of that is the same. Now the fun thing was that which we figured out, this whole space is really. I'm just going to geek out on you because there is no dinner party for where somebody want to talk about scheduling meetings with me in some corner over champagne like I got a PhD in something that nobody want to talk about. But it's really eight little quadrants. So on one side you have internal and external. Then you have 0 on 1, 1 on 1, 1 on n n on n. So 0 on 1 is just Brian adding a note to his calendar of remembering to doing your taxes. So this is you double clicking Google Cal and adding it one on one, you and me.

Host: Friday, April 15th at uh, you know, 1pm I need to do my taxes.

Dennis Mortensen: Yeah. And then there's Brian and Dennis. That's a one on one external as in we don't have access to each other's calendars. Had we worked the same place, it would have been a one on one internal. And I have a shared calendar at least. Then there's one on N M typically small N. That is me doing a demo tool Oracle. There's kind of three folks on the other side. I'll do the demo. I still need to kind of coordinate with the fact that there's three people that need to be available N on N. Brian, Dennis, our uh, college buddies meeting up for drinks, all sorts of kind of complexities. So you really try to solve eight things at the same time. The real not uh, winner, but certainly they extracted a lot of value out of the space, which was calendar. And they attacked pretty much one on one external only through a calendar page.

Host: Right.

Dennis Mortensen: Which is what we've kind of had two major steps in scheduling from a technology point of view. There was a time you will know other people will not know for where we did not have shared calendars as internally.

Host: Not even internally. Yeah.

Dennis Mortensen: Uh, not even internally. Like it just blows people's mind.

Host: Like, but it's like you couldn't see each other's calendars. Nope. No. We all had Outlook and it sucked. And we couldn't see each other's calendar.

Dennis Mortensen: Yeah, that's. That's a relatively new thing. And then I think the next major step was the ability to expose your calendar to strangers, AKA a calendar page of some sort. And Calendly kind of did very well there. But we tried to solve all eight quadrants through an agent, uh, from the get go. When language was not the. Is still not a solved science, but certainly it's way more solved today than it was back then. As in doing predictions in the language space. Today is certainly easier than it was when we got started.

Host: I was at, uh, Conversica, if you remember Conversica, about 10 years ago.

Dennis Mortensen: Oh, yes. Then you have, you have real intimacy with this.

Host: Yeah, yeah. So I, I got the Conversica from rocket fuel 2017 and you know, we just raised a Series B, I think. And they hired me to do all the AI product and there was no AI product. You know, it was all regular expressions under the hood, like bagged words, you know. And so we, we had to redesign the whole system with intents and entities and things like that and conversational brands branching. But it was nothing like what we have today, like building compared.

Dennis Mortensen: So we did exactly what you just described. So we ended up on. I'll be a little bit off, but Directionally correct, I think 47 intents. So that was one of the things I was a little bit.

Host: Yeah, we had a similar number. I think it was about 40, 37 somewhere in there. Intense that.

Dennis Mortensen: So that cover in our opinion. But it turned out to be true to the tune of kind of millions of meetings being scheduled. So we could cover the language universe for meeting scheduling with 47 intents new meeting, reschedule, running late, add participant extend duration. And all of a sudden you think, is this a forever list?

Host: Nope.

Dennis Mortensen: There is actually an end m to it. And we had three entities, right. So temporals, location and people. That's. That's all we labeled against. And we. And this now this is turning into a therapy session for Dennis. So we hand labeled 32 million data points.

Host: We were similar. I don't know if we made it to 32 million. We had a team, uh, we took some of our CS team in their spare time, and we built an internal labeling tool. And so as they weren't working on CS tickets, they would. They would go label.

Dennis Mortensen: We should. We should have drinks one night and just kind of, you know, cry a little bit together. Because we did. The first of all, there was not really any generic annotation tool. So even before you start labeling anything. Well, I guess, yeah. First we built. First we built. We're going on a trip, but first we build a car. Uh, so we had to kind of build the annotation tool.

Host: Like you're trying to build a car, but nobody's invented, like, a distributor yet. So, like, how do you get the power from here to the spark plug? Like, we don't know. Like, okay, well, I guess we got to figure that out, you know, and

Dennis Mortensen: we did that in three steps. So we ended up. Initially, we hired kind of like what you just said. So we hired like, 10 Ivy League kids, smart people, uh, right out of college that would sit right next to the data science team. So that the initial labeling guidelines, which just as an interesting side note, I think we had like, 36 pages just for temporals. Just so you would label temporals correctly with.

Host: Well, then you'd have to. You'd have to triple, uh, up the labels, too.

Dennis Mortensen: Yeah, yeah, yeah.

Host: You couldn't just rely on one human label. You'd have to have three human labels.

Dennis Mortensen: Then you triple it, and then you.

Host: Yeah, I totally remember all this. This is really fun. Really fun stuff.

Dennis Mortensen: We did all of that. And then once we figured out that we are now closer to something where I can navigate this space, then we hired 40 people that wanted to be on Broadway, but are not exactly there just yet. Put them in the next room. Right. Okay. You guys are not tech savvy, but you can read and understand and consume our guidelines and label against it. And that worked. And then after that, we kind of put the whole thing in Manila. And I had hundreds of people in Manila kind of doing labeling. And then on June 3rd, first midnight, uh, Manila time, noon, we stopped labeling. Because the system have now crossed into the territory for the use of the product became the labeling. So big celebration. It was a little kind of.

Host: Yeah, you put the users in the loop as the annotators.

Dennis Mortensen: The very used to the product is you labeling it, which is kind of interesting, but, uh, it was a little Sad because there's 100 people where your job is to do the job so well that eventually I'll have to let you go. But they were, uh. Yeah, they were A good sport about it, but, uh, they were there, all of them, at midnight Manila time, to kind of say, say goodbye.

Host: But, yes, uh, it got acquired in 2021. What are some kind of takeaways from that? From that journey? You worked on this for a long time?

Dennis Mortensen: Half a decade. Yes.

Host: Yeah.

Dennis Mortensen: There's so many good takeaways. I'll give you one good. They're both good takeaways. But one, uh, that kind of worked in my favor and one way I should have seen. Seen the input more clearly faster. So I'll give you the first one for where I was kind of hell bent on somewhat winning the Daily Turing test, because I'm a geek and there's nothing more, uh, exhilarating than seeing the very product that you crafted fool another human into believing that the very thing which they spoke to was not a machine, but some other entity. Yeah, that's fun as a geek, but it doesn't work at all. On all fronts. 1. Now, I can't even upsell them all.

Host: Narrow AI like you cross the scheduling. Turing test.

Dennis Mortensen: Yeah. Like which we. So I could send you so many kind of pictures of kind of chocolate and wine and sandwiches brought to meetings for a machine agent that is never going to be at the meeting itself. Uh, so we have plenty of those pictures in our archive. But it doesn't even work where if Brian used the product and didn't know that this was a machine, you don't even have the idea of. Perhaps I should get myself one of those. Because, no, I don't have 60k, so I can't. I can't hire Amy. So it didn't even work for acquisition purposes. And it worked bad for when things didn't work out, because when people speak to humans, they speak in different ways to the way that they speak to machines. So that was something where I was hell bent and leaned into it like I had Anna, who did folklore at Harvard for ages, to kind of craft the Persona of the agent so that when you spoke to the agent on Friday, it still feels like the same agent on Monday. So really leaned into it. I had a whole kind of team just working on the Persona itself. But I should have seen sooner rather than later that it's okay to be a machine doing machine things in a machine like way, even to the point where no need to write back that Next Wednesday at 1 o' clock is okay for me with Laney. Optional if I know with high probability that that is what this is going to end up being. Just give the man a button man, buttons are not bad. So it took me a little bit.

Host: Like you mentioned, this is what Cal got right, is like we, we attack the external button sharing like narrow use case.

Dennis Mortensen: It was just again, you and me have grown up. We haven't really seen many paradigm shifts on the UI end. So I took my CS degree on the command line. My mom got exposed to compute on the graphical user interface. My kids is probably distinct. Got exposed to a touch interface on that whole mobile kind uh, of generation that is coming alive right now. And then it looked like the conversational ui, which is a distinct UI paradigm in its own right, would be the next thing. And I didn't want to pollute it. But then again, that was not right and I, I, I did see it and I did blend it. And today you can certainly see that on ChatGPT or Pigit, your favorite agent of, uh, the day. Yeah, it's a blended UI experience. It's both conversational on one front, but they're not afraid to throw in a table or throw in a button or throw in visual.

Host: It's multimodal, basically.

Dennis Mortensen: Yeah, exactly. So, so when we started doing that, that worked very well. But it was an interesting thing for where I was hell bent on being in my lane and winning a test nobody asked me to win.

Host: Yeah, I fell in love with the, the product and the solution a little bit. And you kind of drifted away from your, your, hey, focus on the pain here. What is the pain?

Dennis Mortensen: Yeah, exactly. I just, I was just a little detour and then I, then I came back. But, and I remember it because there's certainly a version where, Dennis, if you hadn't come back, you would, I think I would have died. That would have been my, you know, go back learnings from my earlier mistakes where, Dennis, that thing should have turned into a marketplace, but you didn't. And this thing I figured out. Okay, hold on, hold on. Okay, let go there. That's going to be a button here. I think the other thing is, which is a good learning, it just happened so quickly here and sometimes it's, it's much slower and you don't see it happening. But when you do anything new initially, very much so in the B2B space, you are like selling it into an audience that don't even know you exist, might not even know that they need or want what you have. And you do all the kind of, uh, motions around selling, but at some point you cross into a new territory where you are not selling anything. People are buying what you have. And once you arrive in that territory. There's a very big difference in my opinion, between selling and people buying for when you enter, then you need to create an experience where you make it easy for them to buy what you have. You don't need to kind of hyper optimize your selling process now. You need to hyper optimize your dude from the street who already knows this is a pain. You don't need to convince him. He. He'll convince you. He'll have better, better arguments on why it's a pain for him than you will ever have. So just make it very easy for him to buy it. I'm selling. Everybody's fucking buying it now. Like, we had a dramatic growth rate on that. So it just, it was so clear to me where I think I've learned to try to figure out when am I crossing into this new territory, the ultimate territory for any product for where you're not sure if you have multiple vendors in a space, you're competing against them, but you're not selling. The idea of the CRM. The CRM is an idea that's been sold already. So people are buying that. And there's line items and budgets as we speak. Sure, it's now crowded space, you compete on other things, but people are buying it. And that just happened so quickly that I just noticed it so clearly that I've become more cognizant, uh, of, uh, okay, Dennis, when do you need to flip over and stop selling and start creating buying processes?

Host: Yeah, love that. So let's talk really quickly M about what you're working on now. Tell us about launch brightly.

Dennis Mortensen: Yeah, another one of those items on my list. So.

Host: And for those keeping score at home, I mean, we're now like, what is this, your fifth or sixth startup? You've exited three times at this point and you're like, let's do it again.

Dennis Mortensen: This is my sixth venture. We've been very fortunate to have four exits and one that didn't work prior to this. Spent about five years on each, uh, back to back. So in my kind of 26th, 27th year, and we go at it one more time. Incredible side note, I think in many things, if you do it for the love of the sport and not for the trophies, the likelihood of you getting some trophies is higher. But if you're in it for the trophies themselves, you might not get any. And I'm in it for the love of the sport, as in most people, even kind of entrepreneur friends of mine, hate the whole 0 to 1 or the 0 to 10 employees or the 0 to 10 customers or 0 that they don't like. Like just. And uh, and often in the beginning because it's against all odds. Nobody likes me. I got little money, little friends, little customers, little everything.

Host: Yeah.

Dennis Mortensen: Yeah. I find it almost romantic. If anything, that's the part I love the most for where it's probably not going to work. We're trying to, you and me, trying to swim the English Channel. We're probably going to drown, but if we don't, we're going to be on the other side in France. Yes. Like swimming to Cuba from Florida. Like wild. But we did it.

Host: People do that. I didn't. They go from Florida to Cuba.

Dennis Mortensen: Oh, yeah, yeah.

Host: That's. That's amazing. That's even further than. Maybe it's, uh, roughly the same. Anyway.

Dennis Mortensen: Yes, good question. I'll look that up. Yeah, pretty much the same somebody chat

Host: that

Dennis Mortensen: like, hey, I don't like, you know, 3K in the pool. Uh, so. But this is my little replica of that. So I like that. That phase and don't mind. Now, to answer your question, what brought this particular venture about? So, as a founder, your job of the many jobs is to paint a picture of a future that doesn't yet exist. And you paint that to everybody, including your customers. And this is where there's this kind of, uh, odd, sometimes disconnect between your makers and the founders, especially if the founder is not technical and they can add real friction for where, hey, don't go about bullshitting people. I don't like it or it makes me uncomfortable. It's a little easier if you're an engineer yourself for where you can then explain to them why you have to paint a picture of the what is about to happen. First of all, if you're in B2B, our sales cycle might just be seven months. So if I'm about to sell somebody who used this in seven months, I should probably paint a picture of what he will have in seven months, which is not what you're pushing to production today. So on that argument, uh, alone, my job is to paint a picture of what they will be acquiring from us under what particular price point. So don't be uncomfortable. Sure. Be uncomfortable with the fact that we now have to deliver on that, but not on this being some sort of odd lie. No, I, uh, think this is the most honest thing I can do. So they at least know I'm not turning left and the whole product becomes something which they might not think is a good fit for them. So I'm Selling way into the future. Now most of the documentation, collateral, case studies, PDFs and decks that we all have, they're from the past past, the immediate past as in yesterday. But they're never from yesterday, they're from months ago. That means there's some sort of delta, a uh, year long delta between kind of what I am showing people and the world I want them to live in that I didn't like. And certainly as a founder it just hurts uh, to see like uh, this old shit, why are we showing that to people? So the particular pain I wanted to solve was certainly one for how do I make sure that my collateral was by and large if you're a software company is screenshots and similar things of the software product itself. How do I kind of make sure that is up to date or runs along in the same velocity as how quickly engineering can push new features. And here's another side note. We've seen a ton of progression on the engineering side. As in it used to be that we would have launches once a quarter and now of course any junior engineer will put something to production today as we sit here and speak because we can with all sorts of kind of automation and testing and CICD kind of put in place to make that happen. Now as that move forward you just stand still. So I thought we could build if you want to use AI, vocab an agent but it's really automation for where I can log into your piece of software. I can take a screenshot of all your features, I can log into your documentation, Zendesk intercom, help scout, what have you scan all your articles, extract all the images, do a diff, see if anything changed, which it will. If you got any engineering velocity, let you know what changed, give you some workflow for you. You can give me a thumbs up if you like the changes. I'll write it back into the articles and then run that as a process every Sunday night. Not a thousand things, just one thing for where. Well the way we document and expose this piece of software to the world is completely in line with what the actual product looks like when we push to production two days ago. So that's us.

Host: Yeah I'm very familiar with this as a, as a product guy built a lot of product and having the docs team right have, have to be in the loop on a, on a product launch and you know, oh, I need two weeks lead time to, to go through and staging and, and create new screenshots and how to guide. So this is definitely a problem. If you're shipping software every, everybody anybody's shipped software knows through and through, you know, definitely.

Dennis Mortensen: And what I like about this is you can't even cheat. As in, if you cheat, you just pay elsewhere. So it's not like you can say, you know what, we're just going to document half the product and then people will figure out the other half. Now I'll tell you what's going to happen. A whole host of unstructured emails at high cost will arrive in your help center that somebody needs to solve for where? Well, if you told them how they could help themselves, you might have avoided some of them. So you can't even cheat. You must do this. Now. You might not buy our product, but you cannot cheat. You have to do it in some capacity. And it's nice, of course, to be in a market for where any piece of software needs this. Sure, I might not win, I might just die. The next time we speak, my email will bounce and so be it. But it's nice to be in a market where I can call on anybody doing any piece of software. They already have a process doing that and they're already human and Tim would like to avoid it. And I can hopefully come up with something so good they're willing to have Tim do something else and use our product.

Host: That's amazing. So do you think, like, looking forward, you're working on your sixth startup, do you feel like it'll be another five year run and then you'll probably be doing your seventh and then in another five year run you'll be doing your eighth or at what point do you, uh, do you go, you know what, I'm just going to go play pickleball at the, you know, in, in Florida somewhere or something.

Dennis Mortensen: It's a good question. There's obviously people from the outside looking in, painting a picture of. So your plan is to die at the office, Dennis? Because that sounds a little sad to me. Especially because it looks like you have the option at least to not die at the office, but somewhere else. But that also suggests to me that you live a life for where wherever you go during the day is some sort of penalty. And if that is the case, it seems like I'm not the one who should change my life. You are the one who lives a life for where the weekend is what you look forward to. And if not the weekend, then the summer. I just happen to live a life for where me, my laptop and a couple of Red Bulls. I'm living the good life as in this was not a Penalty to get up this morning and kind of start my laptop again. As in time well spent.

Host: I love that takeaway, too. And this is, you know, I, I'm paraphrasing. I heard this 20 years ago with the commencement address Steve Jobs did to Stanford. He kind of tells the three stories in this commencement address. You remember this one? And one of the things he said was. And it really stuck with me because I was working on Wall street, you know, 2005. 6. Living in the Upper east side, riding the 6 train down to downtown.

Dennis Mortensen: Yeah.

Host: You know, touching like 12 people in the subway, just hating my life. Right. Staring at a spreadsheet, you know, 12 hours a day, you know, sneaking out at 9:00 clock to go have dinner with my wife, 9:00 clock at night. And I'm like, there's got to be more to life than this. And, and Steve, I remember watching the commencement address and he's like, if. If I look at myself in the mirror in the morning and I don't want to do what I'm about to do that day, and that happens too many days in a row, I know I gotta make a change. And that just really stuck with me. And, you know, so I just started, like, really trying to live my life like that, you know, like, I really want to do what I'm doing today, and if I don't, I'll. I'll quit and try something else.

Dennis Mortensen: Yeah. Or at least then you try to, uh. I have all the empathy in the world for people for where it's not an option to do today, but then at least try to aspire to what can I then do. To somehow put myself. Uh. Yeah. To put myself in a setting where this is not how I feel. So it will not be tomorrow, but I'm going to work hard to kind of get there. And I think the best analogy I found is that we just. Many professions. But I certainly like the idea that whenever you speak to people in professional sports, they didn't stop because, yeah, I made it. I had that one good contract and you're kind of safe. So no need for me to play soccer or basketball anymore. It was never because of that. Because I actually loved playing the sport. The only reason I'm not playing is that, well, uh, the way it works in my particular profession is that you can't really, after kind of 36, I just happen to be in a profession where I can still pull out this laptop after 36. Right. As in, that's doable. Good on me.

Host: Well, and retired NBA players Don't stop playing basketball. You know, they just play pickup with people their age. Like, I have an NBA player in my neighborhood, actually. You know, and he still plays basketball. Uh, he just plays in private groups with people. You know, he, uh, still has a love for the game. It's not about the money, right? And it can never be about the money. And, and, you know, I asked, I, you know, I asked myself this question when I was working on Wall street, okay, if I had 10 or 20 million in the bank right now, would I go into work today? I was like, no.

Dennis Mortensen: You know, that's the answer.

Host: There's the answer, right? So, like, if, if, if you, if, if you take money off the table, are you still going to do that thing that you're doing? And if the answer is no, then, okay, what would you do? And then start working towards that thing and you, you'll. And I did it many times where I worked towards that thing, and I was like, oh, it's not that either. And that's okay. And then you, and then you pivot and try something else. And, and that. It's. What makes me kind of unique as an investor is like, I've done a little bit of everything. You know, I've had a startup and I've had a small business, and I've done, uh, like every functional area and every kind of company. So, like, I'm kind of a generalist, but, like, I'm not really good at anything.

Dennis Mortensen: But that is probably the profile you would want for somebody on your side of the table, which is, I need somebody who's been exploring plenty parts of the map so that you're not surprised when we enter a new part of the forest. I've been in that part doesn't mean that I'm an expert, but I've certainly been in that part of the forest before versus the, uh. Which is interesting. I wonder if. So, so that, that PhD in some particular strain of some particular thing, who sits at some particular firm, how well do they do over. Fred up at Union Square Ventures, who's

Host: a generalist in praise of the generalist. Yeah, Fred. Fred very much exemplifies that. Right. And yeah, so I think it makes me like, I would have never guessed in my mid-40s I'd be a VC back in, you know, when I was 20 years ago, suffering on the 6 train, you know, would have never guessed it. Right. But it's, you know, another Steve Jobs quote. It's like, easy to connect the dots looking backwards, you know, it's hard it's hard looking forwards, but all you can do is take the next step. Well, this has been the longest podcast I think I've ever recorded. I feel like I could talk to you for another hour or two. I got through maybe 10% of the questions in the outline, so that's a really good sign. Where can folks find. Find you online and and learn more about what you're building and. And you.

Dennis Mortensen: So I'm everywhere on the Internet under Dennis Mortensen. That will be my handle. Feel free to reach out. My email is Dennisaunchbrightly.com if they want to have a look at what we're building. I'm building over@launchbrightly.com trying to, uh, make some nice PNGs. That will be my elevator pitch to non tech savvy folks and off to a good start. So happy to chat on anything on that or anything entrepreneurial. So feel free to reach out.

Host: Very cool. Well, I feel like I just got a masterclass in entrepreneurship. Thanks so much for coming on time well spent.

Dennis Mortensen: Cheers. Fine.

Host: Thank you.