Is Threads The Next Twitter Or Clubhouse 2.0? — With Zach Coelius

28:36They already deployed one. It's called Chat Concierge, and it's simplifying car shopping. Using self-reflection and layered reasoning with live API checks, it doesn't just help buyers find a car they love. It helps schedule a test drive, get pre-approved for financing, and estimate trade in value. Advanced, intuitive, and deployed. That's how they stack. That's technology at Capital One.

29:08And we're back here on Big Technology Podcast with Zach Colias. He's the managing partner at Colias Capital. And in 2020, mid-2020, he tweeted this. The traditional J-curve and early-stage venture is currently non-existent. Usually companies without traction die early and winners take times. Over the last five years, I have invested in 63 companies. So far, only one has been a loss. One, even the lowest performers get funded. There is so much money chasing companies right now. Literally everything is getting funding. Any idiot with a checkbook, myself included in our sector, looks like a genius right now.

29:45We can argue about the cause and the potential outcome, but we can't argue that the beta only has one direction up. So here we are in 2023. The world has changed a little bit. so give us an update yeah it's that we were we were oh man it's so good it was so easy i raised my last fund in 2021 and um like literally it had been like it had been six years of 50 compounding rr every year i mean and i mean i looked like a genius we all looked so smart i wrote my fund it took three weeks to raise in zoom only and never only in zoom no in-person meetings three weeks people were committing on the first call uh it was like i was it was beautiful ah so good yeah we all looked so smart yeah it was great uh yeah it's like wild coyote you know it's like you know it's running running running running running and then just ran off the cliff and now while the coyote is like hanging in midair and we're all like oh how far do we have to fall yeah so uh we had one other company go out of business since then so now i've got two two dead two dead bodies to my record um but we i mean god we're dodging bullets left and right uh i had a company that i was convinced was going to fail last fall that somehow managed to scrape together around and now they're they're doing okay but it was like It was a slog.

31:17I have another company right now that is fighting tooth and nail to stay alive. And it's challenging. And yeah, I mean, literally before the bear market began a year and a half ago, I mean, I would have multiple exits a year and every company would effectively mark up every year. And now it's, I haven't seen a carry check, new carry check in a year and a half. Okay. Being a VC right now in a bear market is a much harder job. It's still a fraction as hard as being an entrepreneur. Like being an entrepreneur is 10 times harder or more. It's so much harder. Your job's harder than being a VC. Like this job is not that hard.

32:09You can make it as hard as you want to. Like you can, like if you want to work 80 hours a week, you can, but you don't have to. Um, uh, and there's diminishing returns, which a lot of people don't talk about in BC, but like, it's still an easy job relative, but it's much harder than it was. Yeah. Back when everything went up. I mean, literally everything, every, yeah. Yeah. It's crazy. What do you mean there's diminishing returns? Oh, so what happens in BC is that, um, about 80 % of that value in the job comes from about 20 % of the work. And then anything you do beyond that, it's kind of like you're pushing a rope.

32:46And that's why you see so many VCs on Twitter and so many VCs writing thought pieces, which are actually garbage and doing stuff that like they're kind of trying to get out in the ecosystem to like raise their name recognition. but it's the value of it is de minimis because the real value comes from people who you've built relationships with over a long period of time that they trust you and they come to you first before they come to anyone else. They're like, Hey, I got this thing. And you're like, great. Um, uh, like, so I'll give you an example. Uh, it'll play into our next segment. Um, so there's an entrepreneur, uh, I had funded, uh, 2017, 16, 2017, 2017, probably.

33:30Uh, and he He built AI for drive-thru restaurants. So, you know, you go to McDonald's and you're like, hey, I want hamburger and fries. And McDonald's bought the company. And it was a great outcome for everyone. We're all super happy. And then so in 2021, he's like, he gets free and he calls me up. He's like, I'm free. And I'm like, I'm in. I mean, my check is here. You just tell me what the price is going to work out to be. And then in November of 21, he comes back to me. He's like, well, I've got good news about this. the good news is we raise a shit ton of money the bad news is the price is crazy high and i'm like you're you're a pro i'm in like i said it was in i'm in um and so then during 2022 i was like god i'm an idiot like i top ticked the market like my the first check from my new fund my lps are like god he's an idiot um and then then the product came out in um 20 uh this this last 2020 the end of 2022 so like you know six months ago and um it's so good it's it's literally lm's generative ai for call centers so you know you call the airline you call hotel and it's better than a human demonstrably better than a human it literally like it speaks whatever language you speak it speaks at whatever speed you speak at it speaks whatever accent you need to speak in and it answers on the first ring so it can go from one to a million agents concurrently so literally you call an airline you're never going to wait on hold ever again waiting on hold will be like dialing a phone like we talk about dialing a phone like most kids don't even understand what that dialing a phone means but like but literally like they're waiting on hold will be in the english language will be just as antiquated as dialing a phone in within a few years um you'll never wait on hold again.

35:19And so it's like, and then the great part is, is like, like that entrepreneur is so good. That company is going to be such a big outcome that like, you know, I didn't have to do a lot of work to get that, just have this relationship that I've had with that guy for over a decade and do the work in that process, um, to be in that position. Um, so it's like one of the great things about VC is that you just, you're, you're effectively harvesting the, the, the labor that's been put in, in the previous years. So one of the, so in the good times, right, you invested in these companies and they kind of, you know, they kept going and now it seems like, tell us a little bit about like what's going on now, like the companies that you funded during the good times, like, are they like, you know, kind of at the end of their runway and trying to raise again and struggling or are they like if they batten down the hatches and laid off a bunch of people to get profitable and survive what's the sort of state of startups right now i mean all of the above the good news is is that like the majority of my portfolio generates a lot of cash we do i think we do north of i think it's 800 million in revenue across all the companies in the portfolio now like it's just like it's a lot of money coming through those companies so like But the vast majority of them, at least from a revenue perspective, are doing fine.

36:43They've just had to batten down the hashes. And they've found that the growth has slowed. And as their growth has slowed, they've had to cut back and stop spending as much. So then their growth has slowed more. So that feedback loop and then all the other companies in the ecosystem are doing the same thing. And a lot of I invest in B2B software companies. And so they largely will often sell to other startups. And so you've seen this sort of like sort of like feedback loops on feedback loops, which is that like companies stop spending on growth. So they stop growing and then they stop spending on software from other people.

37:17So then they stop buying and then vice versa. And so the whole ecosystem has basically deflated as a result of no more free money in the ecosystem. Growth has slowed down. That's, you know, that's normal. That's part of what a bear market is. It weeds out a lot of the companies that were playing company and pretending to be real things. And then it weeds out a lot of these companies, a shit ton of these companies that the VCs were funding with a bunch of dumb money and that were all just fueled by dumb money. And those companies are going out of business. Thankfully, I don't invest in that shit.

37:50So, like, I'm not getting hit yet with that. But the whole ecosystem is getting hit. And so everything is going down. And so, you know, the companies of mine that are, that have not yet found revenue at scale, they're in a tough spot. Like they have to find product market fit or they die. And what about the markdowns? Like, is that also becoming a thing that a lot of the companies that you're working on are doing? We have a company that did a markdown recently. you know that's at the end of the day there's a couple different ways it can work out one way is is you can have what's called a pay-to-play situation the way a pay-to-play works is they go to all the investors and they're like hey guess what you need to invest at this price usually it's a reasonable price you need to invest x amount of money or your holdings will be diluted by 90%.

38:48So before you own 1 % of the business, if you don't invest now in this round, pay to play, you're going to lose 90 % of your investments. So now you're going to own one 10th of 1 % of the business, massive dilution unless you invest. So there's pay to play. I haven't had a pay to play happen yet in my portfolio. Thankfully, they're in good shape, but like I, it could happen. It's possible. Um, and then, then you just have to re underwrite the company. You have to go in, you look at how they're doing. You talk to the, you talk to the customers, you talk to the team, you're like, okay, do I want to re underwrite this company at this price?

39:26Yes or no. It's just a decision that gets made on the spot. Um, and then you have down rounds, which are company goes out and they try to raise money and the money that they're trying to raise is, um, uh, at a lower price than they would have raised before. And the problem with down rounds is that you end up with this situation where the existing investors have an incentive to play chicken with the new investors. So I'll give you an example. We had a down round right at the beginning, actually, of 2022. So right at the beginning of the bear market. And the founder had gone out and tried to raise money and tried to raise money and failed, failed, failed, failed, failed.

40:05And he was on bad terms with his existing investors. It was a weird, weird, weird dynamic. And the new investor comes in. It's like, I want to invest in the company. But as soon as you have bad terms with your existing investors, fuck them. I want them to take a 90 % write down. So a total cram down of existing investors. And the existing investors were like, fuck it. Shut it down. Just shut down the whole business. They're playing chicken because they knew that they could get a better deal than that. that the, and, but, but their leverage was to just shut down the whole business. And so the entrepreneur calls me up and it's like literally crying, like, oh, this is bad.

40:43What the fuck? Like, I'm trying to save the company here. And I'm like, dude, this is a negotiating posture. This is how it's going to work. And then the new investor is like, why am I even bothering with this? Like, what are you doing? And I'm, I'm a reasonably big size shareholder in this company. I, I mean, I had maybe like 500 invested in this company. and so like it's enough that i care so and i call the existing investor how much invested what's that how much did you have invested in 500k okay yeah um so like it's enough that like you don't want to lose that yeah well i mean this is my job right like right so you know the current the current portfolio is about 100 million of deployed capital um so like i have to spend my time accordingly and you you want to spend it on the winners not the losers and so you you know When there's a write-off occurring, that is what it is.

41:32So in this instance, I go to the existing investor. I'm like, hey, there's got to be a way we can work this out. And so this ends up being six weeks of phone calls between these three parties with me being the mediator trying to get this thing sorted out. And of course, it's sorted out exactly where you thought it would, like at a middle ground that's fair that everyone was happy with. And we all took some dilution and the new investor got to dilute us and everything got sorted out. And now the company seems to be doing okay. But like these down rounds are like just incredibly time consuming. Like it's just a real, because this chicken that gets played between the existing investor and the new investor, the leverage that they both have is simply to be a stick in the mud.

42:18And so then you end up with this very complicated dynamic that's very time consuming. and so when that happens, it's painful, very painful. Now, I wonder a little bit about AI because like there's been this moment and we spoke about this with Elliot Brown last week, but there's been this moment where like people say AI and all of a sudden it seems like it's brought back like the roaring, you know, 2020, right? And now it's like, well, how much of this is actually investable? There's been reports that like actually interest in chat GPT is going down. I had it on Google Trends. Some people had it on like the actual web traffic to this stuff.

42:53So I'm kind of curious, like, are you seeing, so you mentioned the call center application. That's very interesting. Are you seeing like real and practical uses of this latest wave of generative AI? And can you share a few examples? Yeah, yeah, no, I mean, I've seen quite a few. I'm a real believer in what I'm seeing. It's not perfect. It's not magic. It It doesn't solve all the world's problems and it won't invent, you know, an endless tub of ice cream that we can always have just sitting in front of us. It just automatically replenishes. Not yet anyway. But it doesn't, it is powerful. So, for instance, I have another company in the business process outsourcing space.

43:35So they basically, they take things that were done by humans in the Philippines and in India before and they automate them. And it's like, it's better. It's demonstrably better than what they were getting before. It's cheaper. What type of processes do they automate? Um, they're not very public, so I'm not gonna, okay. I'm not gonna talk about what. Like I used to report on this company. Oh, what's it called? Names escaping me right now, but it's, uh, they, they, yeah, they did robotic process automation. Is it similar to that type of stuff? Similar in a way. It's different. This is, this is different in its own way, but they're, but yeah, there's, there's some similarities, but, but the end, the end result is, is that AI makes that better.

44:14You know, for the call center thing, AI makes that better. um for i have another company called fireflies you might know they do um uh they do your zoom calls they basically transcribe your zoom calls and they turn them into action items and and i mean fireflies is i mean ai has made they got early access to gbt4 and i mean god that product got it was already good and it got really good like people are if you're not using fireflies right now for your zoom calls you're making a mistake so wait what does it do it just transcribes and then gives you like some notes and action items? Yeah, so it transcribes the entire call, figures out what you discussed and then creates all of your follow-ups and action items and it can push information into your CRM.

44:57Like it does, it's, I mean, if you're a professional and you're doing calls that have anything that you need to keep track of and you're not using Fireflies, you are literally making a mistake. Yeah, the company that I was thinking of again is UiPath. Yeah, UiPath, of course, yeah. Romanian company, amazing company. Great, yeah. yeah, no, no, UiPath is great. But AI has taken that whole market and really dramatically increased the performance of the products. And so if you think about that, at the end of the day, all software products are, they're like tools to help people do their jobs better.

45:32And when you make those tools better, then people can basically get better outcomes and they become more efficient and the world becomes a better place. And so broadly, I think AI is, it's a big, big monster, awesome trend. And we're seeing a lot of consequences. When it comes to the consumer use case, like ChatGPT as a consumer use case, I still prefer to search and read the source materials versus go just ask the question of ChatGPT. I find the answers are kind of weak. They're just sort of soft and they don't have an edge to them that I find interesting. and I don't understand the conflicts between the various different positions on a subject.

46:19It's just like, oh, here's the answer. And then you go read the source material and you're like, well, it might be the answer, but it just seems like it's sort of the mismatch center. Like it's just like, take all this, mash it together and come up with the most probabilistically sort of vanilla answer we can get. So I have not found, ChatTPT has not become like one of my main go-to sources, but that's just me personally. And I'm not a consumer investor, so I don't have a good sense of like where consumers are thinking about it from a personal use case. Yeah. And it's just one of those stories where the data is also just like very instructive, I think, in terms of like, all right, well, less people visiting.

46:53I mean, I know they did release an app, but I don't see it very high in the app store. So where are you on product? I mean, you're doing AI business software. Where are you in terms of productivity and jobs? Do you think that this is going to be something that helps create jobs? or well i mean take the call center one right good example um when that happens the pace that it will happen at will be so fast i i think it will be massively destabilizing like so think about if you're an airline and when when you know when united rolls out you know this company's solution um any other airline that doesn't have it within 12 months is going to be like they're screwed.

47:37I mean, I'm not going to do business with an airline that doesn't answer on the first ring. Like, and like, I get that now because of status. And like, when I have to deal with that airline that doesn't have that, I'm like, ah, I can't handle this. And so now every airline will be able to provide first ring better than a human AI call center. So they're all going to change very rapidly, right? They're all going to switch. And all those jobs are going to disappear overnight. Like, I mean, you're 90 % of those jobs, give or take, will be eliminated. And when you do that across all the call centers in the world, it's very destabilizing.

48:15So what happens to society in that case? I mean, look, long-term, I think the, oh, these tools make us more productive and increases economic benefit and everyone's going to be happy. Short-term, that's the problem, is that like technology now is speeding up the rate of disruption and destabilization. And I don't think people handle that. Humans don't handle these things well. They go fucking crazy. And so like when you take all the call centers in the Philippines and you lay them all off all at once, that's going to be a big problem. Yeah. So when you read Mark Entry since, will AI save the world, I say?

48:53Well, he's got to... At the end of the day, do I want a world where we create things like AI to make us more productive and smarter? Yes. Do I think most of the doomers are idiots? Yes. So I largely align with Marc Andreessen. But at the end of the day, like. There's nuance in the middle. Am I afraid of the consequences of AI in terms of destabilizing human systems? Absolutely. am I afraid of AGI in the long term? Yeah. It's not a zero. Like there's a real risk of like long-term, long-term AI cataclysm. That's, that's, that's not zero. And like two or three years ago, before, you know, we came out with LLMs and generative AI, the number was a fraction of what it is now in terms of like that risk number.

49:47And, you know, on the other hand, humans are just so stupid that like if you leave us to our own devices, like we i'm more scared of humans and nuclear war and global catastrophe as a result of the way that we treat this planet than i am that much more scared of that than i am of of of of agi destroying all the humans so therefore we either we have to we have to become smarter or we're going to destroy ourselves because we're so stupid yeah i'm in the same camp um okay so let's talk about media math media math is one of the most important ad tech companies i think of the last 10 years and they just completely went bankrupt and they're shutting down.

50:25And it's just an interesting story to me because like the way I think about it, if MediaMath can't make it, then nobody can make it. Well, that's not true. TradeDesk is a large public company. But they are the only survivor really from that. Well, there's TradeDesk. I mean, at the end of the day, what happened to my company, Trigit, was the very beginning of the emergence of this sort of like duopoly or triopoly, which is Facebook, Google, and I guess Amazon, they just crushed the open ecosystem. And if you're an advertiser, you have to work with them. And if you were a supplier of tools for advertising on the open ecosystem, you got your ass handed to you.

51:16Trade Desk has done very well to navigate that and they built a big business. but otherwise everyone else is dead so what do you think unless you do you play in the sort of gray areas there's lots of money to be made in the gray areas like if you're if you're comfortable with ethically and morally unclear things and there's plenty of money to be made in those areas yeah so what are the um morally unclear things oh well i mean if you want to advertise products that google won't let you advertise google won't take your money there's people who will take your money. And so if you're willing to use tactics and strategies that Google or Facebook is uncomfortable with to get users to engage with ads, and there's lots of ways to do that.

51:59In the old days, it was pop-ups and pop-owners and it was spam. And there's lots of ways to advertise in ways that are ethically and morally dubious, but that make a lot of money. There's advertisers who will be more than happy to fund those ads. There's lots of ways to make money in the ad ecosystem um just you know are you building equity value in those businesses no but you can put a lot of cash in your pockets yeah i have friends who have big yachts from that shit right what do you think happened with media math in particular was it just that there was only room for one or were there strategic mistakes there uh i i think the um it's it was a challenging it look if the ecosystem had stayed open i think a lot of us would have succeeded and done well and had great outcomes When the ecosystem became closed, the number of chairs in the game of musical chairs evaporated and it became very challenging.

52:54Like, remember, the beginning days of those of us who were the sort of original DSPs. So there was MediaMath, there was Churn, there was DataZoo, there was TradeDesk. And for those who are listening, that's the systems that you would buy ads with. Yeah. So MediaMath is what's known as a demand side platform. It's a tool for advertisers to buy ads out across the ecosystem. And when it first came out back when... So we all kind of built on this concept of real-time bidding being sort of a way to buy and sell advertising. And we all... Some came before that and pivoted into it. Some of us kind of pivoted into that once we saw that.

53:36But that was sort of the gun, the starting gun going off to become sort of the future of the open advertising ecosystem. And we all kind of bet on that. And so in the beginning, there was this group of us who were there. Invite, they were smart, super smart. They got bought right away for a nice outcome. And then they went to go build a multi-billion dollar business. So those guys are great. Awesome. And then of the bunch of us, MediaMath was really the only one left besides TradeDesk. Now, a bunch of new entrants have come in subsequently. And that market has remained a very hot mess in a lot of different ways.

54:14But yeah. So it's largely the, when you say open and close, so it's largely just the big tech companies that came in and closed the ecosystem effectively. We went apps and they figured out how to do the monetization there. And that's what happened. Well, so for instance, with Facebook, before Facebook went public, they missed their Q4 because their ad platform was complete garbage. Like they had, they had designed it wrong. and they, at that point in time, and there's a great book by this, Antonio Garcia Martinez wrote this book, Chaos Monkeys, about this whole journey. And he wrote about, we were in that, about part of this whole process.

54:51He wrote the book. It's a good book. And, but what happened was, is they missed their Q4 because they had built their ad platform wrong. They just didn't, just didn't understand how to build it the way it should have been. And then they, Zuck goes to the ads team and he's like, guys, what are we going to do? And he's like, give me your 10 best ideas. And one of the ideas was to open up the Facebook ecosystem to allow independent companies to manage the advertising and target the advertising on Facebook for advertisers. And so Zuck said to the ad team, great, do all your ideas. And that was one of them.

55:26And so then the ad team came to companies like us and were like, hey, you guys want to buy ads on Facebook? So it's an open ecosystem. Like we were allowed to participate in that ecosystem. And we were like, absolutely. We're there with Bellson. And, you know, my business went, we made it work. Like literally the click-through rates when we were first advertising on Facebook. So before we got in there, they were lucky to get one click for every 3 ,000 ads. So you run 3 ,000 ads, you get one click. and they were making pennies per thousand ads as a result. We were able to 100X that, literally 100 times more efficient and effective.

56:08And so instead of getting one click for every 3 ,000 ads, we were getting 30 clicks for every ad. Like we crushed it because we had built good tools that targeted the right ads, the right people at the right time for things that they wanted. So you go to booking.com and you're looking for a hotel in Sao Paulo. And we were like, hey, here's a hotel in Sao Paulo. Here's the price. Here's the dates you're looking for. And people clicked on those and they bought those. And so the ad prices that we were able to pay Facebook went up by a dramatic number. So now Facebook sees this and their ads start to take off.

56:43Their business starts to take off. And they're like, holy shit, this is how we do it. And so then there's this big meeting that happened between Cheryl and the ads team. I don't think Zuck was there, but there's a big group of them that get together and they're like, okay, do we stay open and let all these independent providers buy media for the advertisers? Or do we fuck all those guys, build the tools ourselves, and close and go direct so the advertisers have to buy from us directly? They chose to go direct and to become a closed ecosystem. So I have a scar across my belly from one side to the other from getting gutted by Facebook as a result of that decision.

57:20And my business, I mean, we literally went from a million in revenue to 30 million in revenue in 12 months. And we had 300 million in revenue on the books. Like we were like, we would have blown it out if they had stayed open, but they went the other direction. They killed it. How was that? And so the whole ecosystem has largely done that. So Google has done that. Facebook has done that. Amazon has done that. They've all kind of become very closed. Okay. Last question for you. Still living in San Francisco. I am currently still in San Francisco. No, Miami plans on the books. Oh no, no. Miami's never been my jam.

57:53um i mean i'm i'm deeply deeply deeply frustrated with this city i'm deeply frustrated by the idiocy that they have driven the city into the ground um it's choppy right now there's there's there's there's hints of of goodness again and it's they're cleaning up they're doing a good job cleaning up parts of it but like i mean they they literally destroyed this city and it's going to take years years to grind our way through this um the big the big question for me is what happens in the next election like if the voters of san francisco are like oh we want more of this i've become very depressed and dispirited if the voters of san francisco are like okay enough of this let's get rid of these idiots that like are just like i mean these these these these there's ideologues run the city and they just do stupid things after stupid thing after stupid thing and And it's like the voters, they put up with more of this.

58:48It's going to be bad. Well, I'll be there in 10 days. So maybe you and I can have a coffee. It would be nice to see you in person. Come have coffee in my hood and I'll show you around. That'd be great. Zach, thanks so much for joining. Great to see you. Oh, my pleasure. Great to see you too. Awesome. Thanks everybody for listening. Thank you, Nate Gwatney, for handling the audio LinkedIn, for having me as part of your podcast network. And once again, to all of you, the listeners, if you've been listening for a while, want to give us a rating on Spotify or Apple podcast, that would be awesome. If you're a new listener and want to hit subscribe, we do these flagship interviews every Wednesday.

59:18And then we break down the news on Friday. On Friday, Ranjan Roy is back with us. We're going to cover everything that's happened this week. You don't want to miss it. All right. Thanks again for listening. We'll see you next time on Big Technology Podcast.