His Public Company Blew Up. He Came Back Solo. | Francis Davidson, Odessia

23:04Francis Davidson:and because we've been using these coding agents so effectively, the amount of time that it took for us to go from idea to reality has been compressed by probably like a factor of 30x given the size of the team. It's been absolutely, absolutely mind-blowing and not like AI slot. There was definitely like a turn in December, January where the models, the coding agents became incredibly good and so we just managed to build without really thinking about the, let's say the lean startup methodology. It was really more about what can we do that's really delightful right from the start, even though it's much more ambitious.

23:38Francis Davidson:So I think coding agents are rewriting the playbook for startups where you used to have like extremely finite resources, finite capital that you can turn into very few features. So you had to have an MVP. And in my view now, actually, you should have something that's like much more extensive. So I think there's probably an opportunity to rewrite the startup playbook at this point in the age of the coding agents. Mike Vernal just wrote something about this recently, about this idea that the wedge is something that maybe companies used to have for a couple of years before they sort of expanded beyond it.

24:10And it kind of sounds a little bit like what you're discussing here, which is like, hey, we could have just started with the hotel product, but we realized we can and we should probably just go more intensively towards the full thing. can you say a little bit more about I'm really curious if anything if anything sparks on that but also you're building a product that has almost this like interesting unpredictability in some ways about it because ultimately so much of the interface is kind of generative based off of the user prompt that it's it would be interesting to actually know how much of that you could actually predict in some ways so for instance you know I'm guessing it was somewhat of a long tail permutation where I was looking at bike touring and I was trying to look at sort of planning a trip around bike touring and it did such a remarkable job of asking the right questions and having sort of the right questions in the UI but at the same time you're able to do that sort of level of sophistication and long tail sort of serviceability that I just couldn't imagine you know in a prior era before AI you could get to in year three or four you know so I don't know Anything that's there that would be worth jumping off on?

25:24Francis Davidson:It's been so fascinating to see how people are using the product. Actually, just last night I was looking at a summary of the last 25 interactions that had happened in the last couple hours. And just the use cases, the kinds of trips that people were planning, the edge cases, it's really phenomenal to see just how different it is. But in the travel industry, we constrain you to a search box. It's like, where are you going? And when are you arriving? And when are you leaving? And then you see some results that are basically generic. oftentimes actually not even generic they're just pay to play i know this from having been on the hotel side of the house you want to get to the top of the search results on booking or on expedia you just pay like it's kind of a true it's it's it's it's sponsored but even if it's not sponsored there are ways to get to the top of the algorithm by being part of their preferred program or by offering them the genius discount say on booking where you're giving a lower price than your own price on your own website cannibalizing your demand in order to get a better ranking on booking so there's all these games um but uh for us it's really it's really actually not that complicated we just let the agent come up with the best possible recommendations fully unbiased so we don't tell like the agent doesn't even know it's like fully double blind in a sense like the agent doesn't know what actually makes money for odysia it only cares about delighting the customer and then it has at its disposal a bunch of tools many of which are generative ui where you can see these beautiful images or you can see i don't know if you saw on a multi-stop trip you can see on the map the various stops with these little postcards and sites along the way and then you see a line and then you see the transit between each of these lines like real time so it's pretty awesome to visualize and you can see multiple iterations of multiple trips you keep scrolling down you see the you can compare and contrast the various ones so we just gave the agent all these primitives that's the kind of stuff that i was waking up in the middle of the night in november december i was like okay like okay it needs to pitch destinations it needs to pitch itineraries and And then this is the way in which it's going to work visually on the map.

27:11Francis Davidson:And so the agent has access to all this toolkit that is very specialized for travel, like dozens of tools that it could use. And then when you compose them, then it could help answer pretty much any travel query in a way that is completely one of one. So it's been really, really fascinating to see how people use it. But it comes down to things that are actually, there are atomic pieces where I think it's a finite list of things that you do when you plan a trip. And we can have all these tools available to the user up front. when you um when you sort of contrast what you're doing with sonder where it was so like intensely i imagine very operationally intensive um with all these physical spaces versus what you're doing now um is there some aspect of like deciding when you were deciding what you were going to work on was sort of that intensive irl like ops off the table was it something that you would just want to try something new were you exploring things there as well i'm curious because sometimes you see people go from pure software to trying to do something maybe more in the physical world or flipping the opposite direction for their second company.

28:16Francis Davidson:Yeah, I think great businesses can be built on both sides. I don't think it's particularly more painful to do something that has heavy operations. It's just entrepreneurship is hard full stop. I'm seeing it now, even the digital product, especially one that has LLMs built in, has this non-deterministic nature to it where it's a little bit like the actual operations of a building where you get some random maintenance issue you'd never heard of that you never thought would be an issue. So I'm not opposed to it. And I think both are equally challenging. What's really cool now is just the degree of ownership that I can have over the product, especially with the coding agents, like the amount of output, the coding output that I put out on a daily basis is just growing.

28:50Francis Davidson:And it's quality and it does work as someone that's like semi-technical. It's been amazing to have such a large impact on the actual experience. Whereas physical operations, okay, we sign a building and well what if it's air conditioning doesn't work all that well we kind of have to deal with it or if there's poor sound isolation and we have to it's it's harder to change versus here it's almost addicting because you can always make the product better you're just like one coding agent like session away from solving an issue if a customer reports a problem or you have this wild idea for something to be magical just go out and build it like the whole product development life cycle of like quarterly planning poker planning just attributing that among a bunch of team members.

29:27Francis Davidson:We're just like a bunch of senior people that are looking at customer use cases, looking at cool things we want to build, and then we're just unleashed independently to go make it happen. What do you think about building a team in this era? Obviously, you have all this incredible experience building Sonder and scaling it. You've learned a lot of lessons through that, probably everything from hiring initial employees in the early days to building out an executive team. I'm curious kind of what you've learned or sort of how you're sort of thinking, how you thought of, I think you have eight people.

29:59Eight people. That's right. How did you go about hiring those people? What were you thinking about when you did? Like, what did you prioritize? Yeah.

30:06Francis Davidson:So I really dislike that organizations tend to scale their output as like a log of their number of people. So I much rather that be linear, if not super linear. Sure. And in order to do that, we need to have essentially parallel threads. There needs to be few dependencies. Like here's one product area you're going to own this week or the next couple of weeks. Everything's tracked in linear and they could just be trusted to run with it. And of course there's some collaboration, some brainstorming that's good. But generally people just spend time in their zone building independently. And that requires not just like decent engineering skills, which is right now it's mostly architectural, I'd say.

30:40Francis Davidson:It's more like an engineering manager reviewing the code of a bunch of more junior, although some of them extremely capable junior AI software engineers. But it's also the product intuition as well. Like I think Mark Andreessen had this comment where he said like there's the 10x engineer, but then if you can do engineering and product, then you're 100xer. And if you can do that in design, maybe you're 1000xer. So there's like an exponential effect to the quality of the output that comes from people that are capable of thinking about multiple parts of the product. So I've tended so far to hire people that are quite senior and can be just given extremely high level instructions and then accomplish something that is that is very high quality in and of themselves and it's interesting because they're not are they software engineers are they product don't it's not clear exactly what that role is but they need to understand software engineering really well in order to be very effective when they came in uh you know this is kind of a new era of building companies when they came and they'd previously worked at other companies would you say that this is like a generally different way of building companies than what they were used to and sort of how did you how did you think about that and sort of setting expectations i'm not not even like it as saying like this is the way it's going to work but more like hey this is like what's what makes it really exciting and makes it different than what you've been doing previously the expectation for how much work could be output in a week makes it impossible not to use like the best possible tools and to use them in a very advanced way plus also we've built our own kind of best practices internally like we have a shared repo for skills, for example, and unlimited token budgets for everyone.

32:11Francis Davidson:We always use the best stuff. And so it takes a few weeks of adaptation. Someone comes in, they're like, whoa, like this is like, how am I expected to deliver so much? And I can't believe you guys do like, I don't know, 20 ,000 lines of code a day. And you merge that into prod and it works. It's crazy. But then we have a, you know, our first founding engineer in particular was like mega bull on this stuff, very early adopter and goes through essentially an onboarding and shows like them, Hey, this is our workflows. And then we share with each other, like what works extremely well um and so we screen for that obviously as well in interviews like one question i ask is you know how like what's your setup to use coding agents really well and some people i've spoken to are like oh yeah just i really watch it do every one of the steps because sometimes it make mistakes i'm like oh so you're only running one coding agent at a time that's that's like and would you be able like open to a world where you were running a dozen and like speaking about a process for that to happen some people are just not trusting of the tools obviously problematic others on the other hand are mind blown at how good they are but also aware of their limitations and that's the sweet spot those are the people in my in my mind that are best set up to succeed would you when you actually like maybe you could talk through kind of like how you ended up hiring the first people like were they people that you worked with previously yeah the first one is like my chief architect from saunder i was like okay let me get someone that's not like a someone who's just like hands on keyboard all the time but could think about systems design and architecture because I think that's the new way of software engineering.

33:30Francis Davidson:So before he was writing and reviewing docs primarily and thinking about architecture, now he does the same thing and it's the coding agents that actually execute underneath. And so I really like that kind of archetype right now. And I think that even more junior software engineers are going to have to get really good at systems thinking and much less like tactical execution in order to succeed. And I think that one thing that's like a little bit different when you're building a company as a solo founder versus having co-founders is, you know, ultimately co-founders, you kind of have like a bit of a culture baked in by the fact that there's already two, three, maybe even four of you if you have four co-founders including yourself.

34:04And then when you add that first hire, then sort of the culture is the co-founders plus that first hire. I feel like in some ways when you're starting as a solo founder, the people that you bring on are so much more influential in the culture because the culture is still like TBD. Can you share a little bit more about sort of how you thought about that and sort of creating culture with the team?

34:25Francis Davidson:Yeah, I mean, I wrote it down from day one. That's something I didn't know. A company needs values and needs a mission statement. Like that we have on day one. It's part of the interview process. Here's the onboarding doc. It's very simple. There's four values. Clean mission statement. And yeah, I mean, I think part of it is that the first, I mean, everyone that we hired until launch, there's the word founding in their title. So I am a solo founder, but I've got a founding team of seven other people. and they have this passion and this vested interest in their success that is very much like a co-founder like I think they're just as excited about about the company and working really really hard and there's a sense of like true end-to-end ownership that's really that's really phenomenal it doesn't go all the way to co-founder in terms of like big strategic discussions like I was alluding to earlier but I feel like especially for this kind of business like I it's very clear that if we have users it's going to work if we have a lot of users it's going to work because you sell a hotel room you make a really good commission out of it that's the that's really the business model simply expressed um and so then it's just about execution so we need a like fewer big strategic debates and conversations we need people that are just bought into the vision and are ready to make it happen so this founding team of like seven folks are just like really um i think uh there's this sense of ownership but it's much more directed towards operationally like making it a reality one thing uh one thing that when we've had other folks on the the podcast in the past they've said is that by having no co-founders the founding team feels much closer to sort of you know like like not co-founders per se but like much more like they're a part a bigger part of it because there's just more space for them i think it's accurate yeah and i and i think that i think there's something really great about that because um you know it just it just allows them to really sort of have this be part of their identity and I think a really constructive and positive way.

36:17Francis Davidson:Absolutely. Can you share a little bit more about sort of like, you talked a little bit about kind of your philosophy around engineering, hiring and things like that, building out the technical team, sort of what you look for there. But what are the things that you're seeing as you've built this like really sophisticated product over the last couple of months? What are some things that you've noticed or learned about building like now with AI that you think are really cool and that kind of surprised you? Yeah. Well, I mean, the charts are unbelievable in terms of the benchmarks, right? And there is, I think it's called the Moravex paradox, like the idea that machines sometimes are much better than humans at certain tasks.

36:55Francis Davidson:And it's easy to extrapolate that across a bunch of other tasks. Like if it can do PhD level mathematics, you know, then surely it could tell if the button is replicated twice, one next to each other and that the front end is broken, but no, it cannot. It's just, it really struggles with certain tasks that like a seven-year-old can do. these coding agents can't actually accomplish, even though they're just like fully in the digital realm. They should have the tools to know. So extraordinary in some regards, but like just the visual perception, design, creativity. I tried to work with all these models, try to figure out a name for the company.

37:26Francis Davidson:Total garbage, can't do it. So it's really important to know like where they're very strong. I think like backend engineering, phenomenal. Prompt engineering, terrible. Really bad at prompting. It's like, that's the one thing I've got the coding editor set up and I'm doing this keystroke by keystroke. The prompt has to be like very, very carefully written. The agents really struggle with that. But overall, I think it's just a matter of time. We're counting the ooms and it's going to get better and better. And I'm not sure that's going to displace like human creativity or capabilities or like that's a whole other separate discussion.

37:57Francis Davidson:But even though the AI can be a tremendous accelerator, there's very fundamental parts of it that it struggles with. Like taste, for example, or coming up with a beautiful brand identity, coming up with a great copy. like it really struggles with these pieces so i'm very far away from thinking that it's possible to build these companies autonomously or that's why i'm not just like the one person billion dollar company kind of idea i think is it would be a very odd outcome in my mind um it would be uh in my view just you probably have to lean on a bunch of like third-party folks that aren't really employees but are helping you do a variety of other things to to actually build a business like by yourself like the the there is something to the an expert coming in in their field and being able to assess the quality of work and contribute that we're very, very far from being able to replace.

38:43I feel like you have two really great names and brands in travel with Sonder and with Odessia now. Can you share a little bit more about sort of how you came up with both?

38:52Francis Davidson:There is no shortcut. It took almost to the T like six weeks for both of them, for the name exercise for both of them. And I tried to hire outside help for both. And in both cases, I'm the one that came up with the ultimate name with the help of others, of course to help kind of direct the ideas but it just requires this obsession it's like you have to be like at hour number 172 and be like i'm gonna keep putting another hour behind this damn naming thing not knowing whether you're ever gonna uncover something better or whether something you've already thought of is the one it's just hiding in plain sight so this one just checked all the boxes for me i could get the dot com i don't like dot ai for a consumer brand i think it's much better to be dot com it's a new it's an invented word type odysia on google first thing that shows up.

39:36Francis Davidson:We have a domain authority of three and we're already at the top of the results. And it maps to the Odyssey, obviously just epic Greek, you know, Homeric epic tale. And I also like the fact that it could be the foundations for a really nice brand identity with like Odessia, which we're making the Greek goddess of travel. This is what our launch video is about. We're like, you know, there is a Greek god of travel. It turns out it's Hermes. Hermes is the Greek god of travel. There's no Greek goddess of travel. We could just make her up. Let's just go make it happen. Create that myth. So we've worked with an agency, a brand agency to help us figure out like the brand identity for sure.

40:12Francis Davidson:But the name in my mind, like I had to have some philosophical connection to travel, had to be easy to pronounce, had to be a new word, had to be beautiful. Like there's so many criteria that get stacked one on top of each other, of the other. And it just takes a really long time to actually arrive at the one. At least that's been my experience twice. And I haven't been able to name something quickly. Did you know when you landed on it that it was right? pretty pretty much off the bat like what it got on the page and you looked at it and maybe it's also a sense of exhaustion after you know six weeks we're just like this is gonna be the one and everyone's like too scared to tell you no and like everyone just runs with it but i'm glad you like it no i mean it's it's really good i mean what i first when you first announced and i saw i saw the video and everything odysia felt just like perfect for a number of reasons the odyssey obviously but i think that the main one is that very often when you and you're sort of alluding to this with like booking.com and sort of things like that um it's usually just sort of kind of point a to point b and that's sort of it you start here and you end there and that's where your hotel is yeah um and the name itself really spoke to this idea that there would potentially be multi point travel yeah um and it's not just okay you're flying to disney world and you're going back home afterwards.

41:26It's you could actually plan a really sophisticated trip. And I don't know, it seems like, you know, in the world where we see more and more sort of things become automated, like these experiences are the things that really matter. And what's actually holding back like these more sophisticated experiences? Well, you said the average trip takes 20 hours, which is both mind boggling to me and also kind of obvious. Like, like, it's like, yeah, it takes a long time, even just figure out what flight you want to take um but when you but then when you experiment with this idea of well maybe i should do a more intensive trip than that and ultimately people end up just giving up because it's like this is way too much to figure out or they happen to be very time rich and they spend their time on it or they get some you know tour that's sort of like pre-packaged yeah and that's probably not the ideal for a lot of people so in many ways i i when i saw odessia i said wow this is something that might actually cause more people to have more substantial trips.

42:22And that's just not something that I really considered, you know, when I thought about, you know, what the, you know, what could actually be possible with something like, something like travel at Odessia. So I'm really curious when you think about kind of what the, like, is that, was that sort of part of it initially? Or like, what was the actual sort of vision behind Odessia when you were getting started?

42:46Francis Davidson:Yeah, there's a very popular trend right now of trying to figure out what are some repeated tasks that people do in front of their computers that we could automate for the enterprise. 95 % of YC companies are just focused on that kind of use case broadly, like AI driven productivity enhancement. But we haven't really spent a lot of time thinking about what are the dreadful tasks that people do, consumers do on their computers, on their phones that we could help make delightful and faster and easier. I think travel is one of those things that has the combination of being beloved by people. Like the minute you have time off suppose you're taking a week off everyone's gonna ask you where are you going like it would be weird if you're just like oh i'm taking two weeks off i'm staying home like of course you're traveling it's not even a question it's like this natural impulse that we have on the other hand very complicated very difficult to do if you're traveling internationally it turns out it's something like 40 hours to plan a trip the vast majority of which is not fun it's like train schedules and it's figuring out like the logistics of it and reading reviews for hotels like oh yeah i heard there's not great ac in europe like and it's going to be what's going to be the temperature again in Rome and you look that up that's another tab like Odessia is just like right there in front of you all the entire trip on one tab and significantly less time so we just started getting excited about the amount of time you could save also remember big student of Jeff Bezos he was an investor in Sonder actually and he often thinks about the value created he gets the question what is the you know is it fair that you are so wealthy off of Amazon he's like well first I mean I own only a small percentage of it the vast majority of the wealth has been created for others, but also think about the amount of time it takes to go to the store.

44:15Francis Davidson:Look at how many customers we have. And every time they push that button, it saves them maybe an hour round trip. And so this is the aggregate amount of hours that I've saved for people. So businesses, you know, the consumer surplus that they could deliver by saving people time is massive. And I think travel is something we're probably going to do more of. It's growing faster than the rest of the economy and it's experiential. It's not good. There's this, obviously this mega trend behind that. Whether you think we're going to have more or less free time in the post AGI world, like I I think it's pretty clear that people are probably going to want to travel more.

44:43Francis Davidson:So there's these really nice tailwinds that exist here. And it's quite complex. And it's even more complex, by the way, if you're a travel hacker and you're trying to optimize. Like that's just your regular trip plan. But if you want to figure out whether you should use points to book your flight or not, and whether actually instead of doing it, these three cities in that order, what if you change the order? Because the dates will have an influence on the price per night of the hotel. And what if it turns out there's a conference happening? You need to check all the permutations. And if you're flexible and you can travel either all of July or August, my God, is that a minefield of optimization opportunities?

45:13Francis Davidson:Because you could just plan the trip for every single one of these start dates for every single one of the iterations of the itineraries to identify maybe a trip that's 75 % less expensive or something I did with Odessa. I was like, Hey, I want to go to Vancouver in the next month. Odessa goes, checks the flights for every one of the next four weekends. I want to just do a weekend trip. There's not a single search box in the internet where you can be checked the four weekends in a row. Odessa does it. And then here's the five hotels I really like in Vancouver, checks the price for all five and then gives me a table oh in two weekends you can get a direct flight you can do it business class you can use points and here's the this hotel happens to be abnormally cheap it's going to cost you like 75 less than the average of the other weekends and all you have to do is push one button to go so it's just it feels so delightful these these kinds of things um and i think a lot of people actually take a lot more trips if there wasn't this activation energy because sometimes maybe like you have some points that are expiring or you have a credit on an airline that has a one-year expiry and it doesn't even cost money.

46:08Francis Davidson:You just need to have that information available and then constantly run some of these searches, these planning sessions in the background to surface them to users. So I actually think a use case that we're going to have at Odessia is going to be even more prominent than people just going on our site or on our app to plan and book a trip, is that we're going to know who you are and what you like, what are your loyalty programs, all that kind of stuff, and just constantly be on the lookout for the best possible trips. So the first thing that you get when you create an account on Odessia is what's on your bucket list.

46:34Francis Davidson:And the reason why we asked that question, because we want to make sure you get to all these bucket list items in your lifetime. And we're going to find the time that works perfectly where it's convenient, where there is a good time of year to go. You're not going to go during monsoon season and the prices are really good. So you can go have better trips for a lot less money. Uh, so I think that's just a whole bunch of things that I think the average consumer would, would resonate, would resonate with. Um, and that I think our team is really excited about. it's really incredible how little context most travel sites and products have on what people actually want um even even when they're actively searching you know that that idea of you just put in the location that you're going to that's about it right versus um when you're putting aside the onboarding for odesio which is really cool just like when you're actually doing sort of the initial prompt for a new chat or a new trip um you have this free form entry you know and that really opens up a lot right because i was able to and i use i often use uh you know voice to text um and that just allowed me to talk for maybe 30 seconds the text shows up in there then things start going and you have so much more information from me in 30 seconds than booking.com or airbnb has and i think that's really powerful the other thing you said that really resonated and um i think is something that maybe people should think about way more is this idea of and you you also brought up bezos you know sure he saved people an hour of time at a store for shopping but that also probably makes more people buy things right because they're like okay well i'm not going to have to take an hour to go to the store changes their behavior entirely you running all these different uh sort of like background sort of searches or or or your processes to determine things and uncover things for people who are on odysia probably means that you're going to expand the amount of times that people travel like significantly you want to plan a group trip you just put together the itinerary in odysia share it so you can have a public link for it send that to your friends see who likes it they could just hop on binary like yay or nay onto that trip like just reducing the activation energy i think is going to make people go out more which i would love nothing more than people to travel more and to like maybe you know spend less time on their phone.

48:48It's really, it's really cool. Um, what, like what's next? What's the, what is the next sort of the next few months look like for you all? Like, I know you just, you just announced, I think it was like less than a week ago. Uh, and thank you for coming on and talking with us, but yeah, well, I'll ask you maybe one or two more questions, uh, before the, uh, before we wrap, but yeah, what's next right now? Yeah.

49:06Francis Davidson:So we just raced to build a good product. Now we've got a bunch of users that are, there's actual repeated patterns of behavior where we can start seeing, okay, it turns out like we have this amazing hotel search, amazing flight search. Um, but people spend actually the majority of their time on the platform, figuring out what they're going to do. It sounds obvious, right? Like which restaurant am I going to go to? What places I visit? If I'm in Tahoe, which hikes am I going to do? And you know, just putting together the whole plan. So we're building a lot of tools to make that better, easier.

49:34Francis Davidson:I think multiplayer as well, just sharing with others, having multiple versions of trips side by side. What if I will actually use Odessia to plan three different versions of it and then compare the cost. and then after that pick the one I want right now we can just only plan one one at a time in a given in a given session so we're getting really good ideas and feedback from the usage pattern so we're gonna we're gonna nail that and then after that I think you know building much more of the background agents to start proactively planning trips for people I think that's going to be pretty pretty magical so I'd say my short answer is we're at the stage now where we're just building in public we decide to launch fast not everything's perfect it's still in preview and we're just going to make sure that like it is absolutely delightful and that people share it with others and that there is just inherent love for the product.

50:13Francis Davidson:And then, you know, step after that is just like the growth engine. That's an entirely separate business and we're only getting started like to think about that lever. But in our case, given that we can build so fast, we're just going to keep investing and just making the product absolutely killer. Now we have thankfully enough users that we get like everyday sample size on every change that we make. And that's the perfect kind of launching point to inform our product roadmap at this point. Super cool. Okay, so last couple of questions here. I'd be really curious you know you had a crazy journey with Sonder a lot of it was due to things that were completely outside of the company's control like with the pandemic and I think all of your competitors died during the pandemic you're the only company that actually that actually made it through in sort of your category what are some things that you kind of learned from Sonder that were out that were inside of your control or at least maybe are going to be inside of your control now with this new company that you think like looking back on or things that you should, you should, you know, like maybe relate to other people that they could maybe keep in mind.

51:17Francis Davidson:And I really love Saunter. It was a great business. I think it could, it could have done phenomenally well if, if, if COVID hadn't like kind of nuked our headquarters at the worst possible time. I really am fundamentally a believer that the story would have, would have turned out great. Um, but, um, yeah, I think, uh, one of the, one of the pieces of advice you often get is that you should hire exceptional talent and kind of delegate and let them do their thing. My experience has been that 90 % of the time that has been an unsuccessful strategy, especially when it comes to executive talent in the context of a larger company.

51:49Francis Davidson:And that the kind of Chesky-style founder mode where you're into the details and you're kind of micromanaging, for lack of a better word, that is a far better way to ensure success. I think that there are a series of disincentives that pop up, empire building being one, maybe lack of skin in the game. Maybe there's like a selection bias as well. People that end up running large organizations and are big executives tend to be further away from the actual building itself and really good at getting promoted and accumulating power within organizations. So I've become a little bit skeptical on executives.

52:24Francis Davidson:And now what I want is just like the, for sure leadership is important and we're gonna grow the team and there's gonna be management layers at some point. But just this like obsession with personal output and being actually a contributor and subject matter expert. I was looking also at an old Steve Jobs interviewer is like thinking about how to hire managers. And he's like, don't hire people that are just managing people. You need to hire the best practitioners because no one's going to want to work for someone that can't actually teach the craft. And so I think that's something that's going to remain the case is like, everyone's going to have an IC workload effectively.

52:55Francis Davidson:And management is something that gets done on the side. It's almost like team leaders is the way to go. And especially with coding agents and the amount of, or not just beyond coding, just agents full stop. You can, you can output a lot of personal work, um, in a much smaller amount of time. And I really also respect the idea that some, some entrepreneurs have done a really good job building talent from within. And I hired too quickly and too externally, I think at Sonder, I think it would be best to just take your rockstar ICs and then give them the opportunity to grow as leaders. And some of these companies have VPs, C levels that like run massive teams, very important parts of the organization that are like in their 20s.

53:33Francis Davidson:And I think that's phenomenal. I would love to have people that are just super high potential, that are experts at their craft, that are also leading departments. Well, you know, the last two questions are the customary questions we ask. I think it's really cool because you have had like, you know, obviously things were challenging at the end of Sonder, but you had really great experience with your co-founders starting the company, and now you're building solo. So we usually ask for sort of the bear case for being a solo founder um and maybe you could maybe you can both apply that from the perspective of maybe like a first-time founder and also if somebody who's maybe built a company before and then after that we'll chop on that for a little bit and then we'll we'll talk about sort of the the bull case for solo founding yeah bear case i think it's only one argument for me personally which is could it go faster like if i had like absolutely world-class co-founders that we're grinding 16 hours, seven days a week alongside me, would we go faster?

54:30Francis Davidson:That's always the question. It's hard to say. I mean, because also sometimes that means things go slower because there's more debate about things. Whereas sometimes having a clear direction, like there is usually one general on the battlefield. And if there's two generals or three or four, like things can get out of hand. Like organizations typically are led by one person. Like hierarchy potentially is good. A dictatorship in the context of an organization could actually be valuable, but also like helping hands and people that are very invested. And if you could divvy up work correctly, then maybe that that's, that's a preferable path.

55:00Um, that's my thought on the first piece.

55:03Francis Davidson:I don't know where else you want to go next. No, I mean, I think that's, I think that's great. Um, I think that makes a lot of sense too, where if you have, if you have the person like that's the right fit for the business, it's really worth strongly considering having that right now. I think that maybe the, maybe we'll go into the the bowl case but I know part of it was about was about this idea that hey if you have this idea this thing you're working on and it's a now thing you shouldn't necessarily wait to find a co-founder you should go after it but what's the in your opinion what's the what's the bowl case for solo founding yeah I think there are amazing people that you can bring in the organization after having founded the company and they could be incredibly talented it's not needed to have like the whole bench up front so actually part of my thought was okay you know let's just build out this product put it on the market and then maybe we like are a little bit more on people's radars and it becomes significantly easier to attract great talent so instead of for example we don't have a CTO right now an amazing chief product officer we have great founding engineers there's no CTO it's fine not to have a CTO for a handful of months and like what kind of CTO can I get say a year from now suppose that we're very like we're taken off 12 months from now i'm able to get the best like extremely high caliber people and actually probably would have wished that i waited that long instead of getting maybe someone that's a little less like you know of a perfect fit earlier on so it's important to be mindful of like how attractive you are in the marketplace if you're a pre-launch company like you can have a lot of salesmanship but it takes a lot of effort that's another thing i was spending a lot of time on i was trying to convince people to join the company pre-launch i was co-founder dating and like the amount of energy it took because if someone's going to be a co-founder how much time do I actually need to spend with them just figuring out whether we jam well together and whether the idea is good and then economic splits and like I went down that path with a few people I was like oh my god this is like an entire 40 hours have gone by and I'm back to square one because it didn't quite work out and like how I'm gonna to do a really really thorough job would take a lot of time like I don't have time for that I just need to go build and especially given like we said I can as an individual now have a disproportionate impact given the the coding agents and just the AI automation that is now possible that every single hour I don't spend just building the product and just speaking with people I feel like it's there's an enormous opportunity cost behind it with Sonder like I could do a lot of work as an IC but primarily my role was hiring people and then directing and then this was the managerial thing being in meetings all day is what I did whereas now I dread a day where I have more than a few meetings because like I need like I could just there was 20 ,000 lines of like you know code that were output yesterday.

57:40Francis Davidson:And thank God there was no meetings on Sunday because I could just get a bunch of the features that are amazing that are delight people out the door. It's amazing. Well, thank you so much for doing this. This is great. I think that Odessia is going to be one of the most important companies in the world. It means a lot. I think that it's going to change the way that people travel, like seriously. I think there's just something really special about this idea that like you're actually going to increase the ability for people to actually go out there and experience the world. I think that's really important, especially now.

58:10Francis Davidson:Amen. Music to my ears. Love to hear it.

58:35YouTube channel, anything helps. And of course, if you are starting a company or considering starting a company, you should go look at the Solo Founders program. We work together with you in San Francisco alongside of a bunch of other Solo Founders. The idea is it's much better to be Solo together than to be Solo alone. I'll spend a lot of time helping you. And we also invest$100 ,000 in your company. So if you're excited to learn more about that, go check out solofounders.com slash program.