TWiST News: Subscription Economics, Density's Waffle, and The Cost of AV Fleets | E2027

0:00This Week in Startups is a show about startups and technology and that journey of trying to find product market fit and solve really challenging problems in the world. This Week in Startups is brought to you by Fundrise. Fundrise provides access to diversified portfolios of private real estate to all investors with their industry-leading, easy-to-use platform. Sign up today at fundrise.com slash twist. Brave. Brave is an internet privacy company on a mission to protect your personal info online. Download Brave today at brave.com slash twist to browse faster, search privately, and so much more, all in a single click.

0:42And OpenPhone. Create business phone numbers for you and your team that work through an app on your smartphone or desktop. Twist listeners can get an extra 20 % off any plan for your first six months at openphone.com slash twist. I have a co-host here. His name is Alex Wilhelm, and he has an incredible 15, 20-year resume of working at places you've heard of, like TechCrunch, Crunchbase. And Cautious Optimism is his newsletter that you can subscribe to for about a hundy a year. How are you doing today, Alex? I'm doing fantastic. If you have young children, you'll know the feeling when you wake up and you're like, wait, I'm not exhausted.

1:23And I must have slept well last night. So I am flying today. And we have a really great show. I'm actually really excited about this. Jason, I'm going to tell people what's coming up. So First of all, new subscription rules. Ladies and gentlemen, for the first time in twist history, we are going to have Jason Calacanis say nice things about Lina Khan. So if you want to say history made, stay tuned. Then we have the CEO of Density On. They have a new thing called a waffle, but no, you can't eat it. Just wait. Lots of cool stuff to come there. We have a first look at our AV fleet financial model, then nuclear power rising and the startups looking to make solar possible jason we are loaded today but first up lena khan the ftc has some new rules out and this caught your eye yeah you know uh we've been very critical about lena khan i did a meme today from one of my uh i have an army of meme uh builders for me now so if you pull up my twitter you'll see this meme so it's the uh vanos meme did lena khan get us click to cancel yes what to cost no exit market for four years so this is somebody clipped this and said it's a lena con thank you lena con for your service we do like the ability to cancel our gym membership but please please lena con if you get another four years before you go to andreason horowitz or y combinator as a partner and secure the bag please lena con let us sell companies for under a hundred million dollars without any approval process please just singles we won't even ask for a double a double would be 500 million.

2:55Okay, I'm going to ask for the double. Anything under a billion, just let us sell it. Under a billy makes no difference. Let us sell. You just went from 100 million to a billion in like two sentences. A billy. A billy. I don't disagree. I actually think a billion would be a perfectly fine threshold. Yeah, have at it. Here's the thing though. We are talking about this new FTC rulemaking in terms of Jason and Alec want to cancel the Wall Street Journal sometimes and their gen. Here's the thing though. What about businesses that sell to consumers that currently make, let's say a good chunk of their revenue, Jason, by folks for getting their pain,$5 a month.

3:26So there's a pro-business, pro-consumer kind of tension here. And I'm curious how you think it's going to net out for businesses. Okay. So I think one of the great things about subscription businesses is that it creates a North Star. When somebody unsubscribes, they get that box. Why are you unsubscribing? And you get information back. I don't live in your city and I moved to, you know new england somewhere i'm in boston i'm in cape cod wherever you are i'm in brooklyn well that's giving you a signal if a hundred people move to brooklyn that's where you should open your next location it's short-term gain as an entrepreneur to not let people unsubscribe and you don't want people locked into a product that doesn't work for them and there is a counter to this which is you can come up with different subscription terms i have no problem paying a year in advance and the person saying you can't cancel you're buying it for the year like my ski pass when i just bought my epic passes and last year i bought an icon pass because i thought i was going to go icon and my schedule changed and i couldn't go on the icon trip and i had bought a 1100 icon pass i skied zero days on it but on my epic pass i had skied 25 or 30 days so i felt fine you know i put the two passes together divided the number of 2 000 by whatever number of days and it was half the price of if i had bought lift tickets per day so if you want that reservation added a nice hotel there's a cancellation period right if you're booking a hotel during the super bowl you can't cancel it it's the super bowl uh or you can't cancel a week out so there's reasonable things to be had here and if you're a sas company and it requires onboarding you could say you can cancel any quarter you know or with 90 days notice in year two just make the just make it up front right is i guess the goal and this creates a weird tension where you have a group of people inside companies, Alex, it's such a great question, trying to do what are called dark patterns.

5:22And dark patterns, no bueno. Because now you've got really smart people trying to figure out how to trick people. That's not how business should work. You don't want people tricked into using your services. It's like you don't want people tricked into being in a relationship with you. This goes for, you know, the ultimate dark pattern in the United States is employment and healthcare. If you are working at a company and you've got kids and you've got healthcare and they've got a good family plan, you stay at that company because of the healthcare. People will stay at IBM or Google or Microsoft because of the healthcare.

5:55And if the healthcare was portable or was provided by the government or independent of employment, or you picked it and your company gave you a contribution to it, which is how I think it should work, man, things would be so much better. So much better. You came here first, everybody. Welcome to This Week in Startups, where we're in favor of Lena Khan and Universal Healthcare is on the table. All right. I'm okay with it. I think we should have a basic healthcare system just for the basics that everybody gets. And then you should pay for on top of that. And you should always be a fee. So if you go in, you should always pay 25 % or 20 % of the cost so that you know the price.

6:31Yeah. That's, I think, would be reasonable. Just for things like basic checkups, that would save us so much money later on. But I do want to give a plug for myself, which is that while we're talking about smaller deals, acquisitions and startup M &A, I do have the CEO and co-founder of Cashflow coming on Twist this Friday. They just sold the HubSpot early stage company. I'm going to figure out how the hell they got that done and if there was any regulatory pressures, Jason. So more coming on this topic. All right. Venture capital is one of the most lucrative asset classes in the world. Don't I know it?

7:04I do it for a living. Why is venture capital so great? Well, just look at today's public tech companies. Almost all of them raised venture capital when they were a private company. And that's great for venture capitalists like myself and for my LPs. Awesome. But what if you don't have enough capital to back a venture fund? Or maybe you don't have an in with the top performers in venture. How can you get the same deals that institutional investors like endowments and sovereign wealth funds get into? Well, unless you know a guy who knows a guy who knows a gal, you aren't going to be able to buy shares in that hot AI startup until they go public when all the value's been created.

7:41But there's another option. The Fundrise Innovation Fund is a$125 million fund that holds shares in some of the most exciting free IPO tech companies. And it's designed specifically for individual investors. Now you can get in early at fundrise.com slash twist. That's fundrise.com slash twist. carefully consider the investment material before investing, including objectives, risks, charges, and expenses. This and other information can be found in the Innovation Funds Perspectives at Fundrise.com slash innovation. This is a paid advertisement. Okay, awesome. Well, we have a great guest today, and he's got a new product, and I think it's time to transition to that.

8:22Yeah, Alex? Yeah, let's do it. Okay. So I used to run a conference called Launch Festival, and before that it was called TechCrunch 50. And I might bring Launch Festival back. We stopped it during COVID, but at this event, 50 people launched a new product. One of those new products was a people counter and it was at Phil's Coffee in San Francisco and it had a laser at the door. And when you broke the laser, it counted the number of people coming in or out and it had to kind of figure that out. And then there was also one that used wifi, which was a little creepy because you could identify people and you would know that the same person came in connected to the wifi because even if you don't log into a Wi-Fi, your phone might connect to it.

9:00That founder is Andrew Farah, and I am on the board of his company, Density. And I was, I believe, the first investor in the company. Welcome to the program. Andrew has a big product announcement today. Hey, happy to be here. Andrew, meet Alex. I don't know if you two have met before. No, actually, I think this is the first time. Although, since we've met, we've talked about skiing half the time. So now, later on, we need to talk about snowboarding. Okay, here we go. Same thing. Sounds good. Andrew, we have been on this journey with Density, and the domain name is density.io, I believe still. And you can go see this incredible software.

9:34And tell everybody what the mission of Density is today, and then let's go right into the announcement. We won't prolong this yet. I'd really like to know how the world is used. And I think there are a lot of other people that would too. There's about 3 trillion square feet. of physical space in the world, residential and commercial. And we don't know how any of it's used. Something like 99 % of it is unmeasured. So today we work with predominantly like large workplaces, large corporate offices, and we deploy radar sensors, kind of like wireless access points. So like above, you know, on the ceilings.

10:13And then we get a heat map of use inside buildings. And so we've spent a lot of time, about 10 years, trying to count people in physical space. and uh that's sort of our core customer is um large corporate offices and then we just announced something yesterday um called waffle uh that has been you know sort of like part i described it as part love letter to our customers and part return to our roots um because it's something that we wished we could have bought when we started the company and we just couldn't we like couldn't find in the world. The technology wasn't there. And for people who don't know, the way density works today is you have to put a sensor in each of the conference rooms.

10:58I'm in a conference room here at Capital Factory in Austin, who is nice enough to house us while I look for our permanent office here. Shout out to my friend, Josh. And if you're a startup, Capital Factory is a great place to hang out. I'm in a conference room with a producer. This conference room has seating for 10. and for a work uh co-working space or incubator like capital factory it would be great for them to know the utilization of different rooms what people do is they use conference rooms for 10 typically as phone booths and so you know with density if you were amazon or google or microsoft or linkedin and you have all this office space wouldn't it be nice to know the utilization level and of the conferences maybe you find out hey we've got too many conference rooms and not enough phone booths, let's convert these two offices into eight phone booths, and that would be a better use of space.

11:48So space planning for organizations, space is typically the second or third line item. Is that right, Andrew? Yeah, second effort payroll. In fact, over the last year, something like 53 % of all meeting room use was by a single person, regardless of size. Tracks, yeah. So those have a limitation. When you put the sensor arrays that Density currently uses, it requires power over Ethernet, I believe. And you have to put these door sensors in. And I don't know the approximate cost, but maybe you could give people an idea of what that costs. And maybe you could show that graph. I'll hold$1 ,000 all in.

12:25Installation, hardware, software. And something like between 50 % and 65 % of that cost is installation. Got it. So it is hard and it costs money. But once you have it up and running, now you have perfect visibility into your campus and if you had a lot of space if you could cut 10 of the space or one percent of the space i would assume uh sublet one percent consolidate not add space it would pay for itself yeah oh yeah i mean to give you a sense of scale like you know we work with many of the fortune 500 fortune 100 they'll have 65 million square feet of office space of office space not not warehousing not like other sort of medical or infrastructure we're just talking offices um and when you're in the tens of millions of square feet of space even north of a million square feet you just lose visibility like so you just like can't you can't know you can't put eyes on all the space the pandemic put a very like sharp point on how limited our visibility was in buildings because everybody went home andrew so when i'm thinking about this though, will I put the new product, the waffle into each one of the rooms at my Fortune 500 company office footprint?

13:40Or do you think that the new product is more aimed at companies that have, I don't know, a couple hundred employees? I'm trying to sort it out and kind of put it into a market category, if that makes sense. Yeah. I mean, we were solving a problem that we faced every day doing deployments, which is like, it sucks to spend$700 on installation and$300 on the technology. That sucks. The question was two years ago, can you delete the installation cost? So we deleted the installation cost. You now self-install the unit on a wall. So you went into founder mode. Let's call it what it is. I was there. I saw it.

14:18It's been a rough couple of years. Eating beef and rice and going to the gym and then working on density. That's essentially all that I've been doing. Founder mode. Congratulations. Let's show the product. And I love this chart talking about what you did in terms of what Elon does in founder mode or Brian Chesney from Airbnb. So here's waffle. Gorgeous. It sits on a table. I guess it gets attached to the wall. It looks like a hockey puck or a waffle. And I suppose you could put this into an Airbnb as well to know if they're partying, huh? I mean, sure. I mean, I think the thing that's exciting about self-installation is that it immediately addresses the biggest pain point our primary customers have, which is cost.

15:00But it also opens up the world to lots of creative uses for count inside space. Got it. Um, yeah. So I, I'm excited to, to make this extensible to developers, especially, and also folks that are just interested in like, if this, then that style things, because I think the uses are, are, there's a very long tail. Oh, that's a lot of fun. So if you did, if this, then that, you could say, if there's more than five people in the room, turn the walls purple. And then if there's less than three people, turn them blue. And then you can have the DJ know how many people music. Oh, that's actually really fun.

15:32I was thinking totally boring. I will say it is capped count. So, so wall mounted self-installed units have an occlusion problem because, you know, they can't see the entire space. And so at launch, the product will do what we call capped count, which is zero, one, two, and three plus. And then over time, we'll sort of iterate on the algorithm and kind of update it, but it gets to most of the use case of is a room in use and how yeah and here's um really where this gets super powerful um people don't get to see this all that often but here is an example of a building and here's floor number five office 403 and what you'll see is you can see that basically a heat map at different times and when you're just looking at the floor plan even if it was desks or you know um office space conference rooms you just see the density of the space and this gives a facilities person or somebody who's doing return to office and wants to know if there's asses and seats uh if it's working or not right because now you have amazon saying hey everybody's got to be in the office andrew jesse said uh apple's doing three days a week people are starting to come back to the office and realizing the value of it including myself and so here we go you're getting to see the rooms light up.

16:51So when you work with - Can I mention what this is? Can I mention what this is? This is live wayfinding. This is for occupants. This is for employees. Got it. It's on screen over by the elevators where it says you are here. Mm-hmm. And this is real data. And the way it works is we have a grid that shows the major highways of the floor plan. So this is ways for your office. That's right. It's ways for buildings. Ways for buildings. And so the sort of aquamarine color is in real time available. And the red is currently occupied in real time. And real time is defined as essentially sub 700 milliseconds.

17:32So if you walk into a room, it will instantly turn red. So this is fantastic because one of the big complaints people have with office suiting as a concept or conference rooms is, oh my god i got there somebody was in the room who didn't have the room and not only were they in the room they were in the room with a bunch of chinese food and eight people and then i have the room reserved and i gotta tell them to clean up all the mugul guy pan and the general so's chicken the sauces everywhere it's a disaster i mean we've all had this experience i mean we should have called the product mugul guy pan i think i mean i'm just rattling off dishes here i like but uh yeah it's such an incredible product and so this you had a chart i thought that would be very interesting to show you this took four months to deploy oh what you're looking at right now took four months to deploy all the devices all the planning all the poe all the ethernet cables like all the ladders that you had to climb up on after hours because you can't do work when people are in office this took four months to deploy with waffle um you could do this in an afternoon yeah wow you could do it literally in like two hours do all this is how many sensors How many waffles will we need, Andrew, to make this work today?

18:45One per room. One per room. So there's one, two, three, four, five, six, seven, eight. I don't know, 15 rooms here. So call it 15, six, 20, 20. Is that a reasonable estimate? Yeah. Okay. So, you know, somewhere around there. Yeah. 150 bucks a piece, eight bucks a month. Seems pretty cheap. I mean, when we think about the cost of office space, eight bucks per month per sensor, which I think is what you're charging for waffle. Doesn't seem like a lot, frankly. I almost wonder if you're undercharging. He's undercharging. That's my advice. as a venture capitalist. Jason wants you to add a zero to that if you don't mind.

19:16No, I mean, one of the great things, Andrew, I'll let you answer the question is when you underprice something, people will use it more and they'll deploy it and get value from it and you can scale. Yeah. You remove an objection because I could see people saying, oh, a thousand dollars a room to install these 40 ,000 per floor, 10 floors,$400 ,000. That could be a blocker for somebody who's got a company that's really penny pinching or consolidating at that time, even though they would save it in the next year with what they would learn, right? I mean, this utilization, the sensing of buildings, like buildings becoming aware of humans, I think is going to follow a similar adoption curve as Wi-Fi.

19:59Wireless, buildings 30 years ago did not have wireless connectivity. They were all hardlined. Yeah. And if you look at the history of AlohaNet and all of these, it's a very rich history to Wi-Fi. It wasn't until 2011 when we hit 450 megs down per second that it was like, this is as good as hardline. I should just have that. This is silly. And I'm now talking to you on a laptop as a result of that infrastructure being deployed. But it took literally 25 years. And so I think sensing is kind of similar. We're not replacing an existing system. It's net new infrastructure for buildings. And so if you get that cost curve, if you're on the bottom end of that cost curve, you can rapidly accelerate the adoption of new infrastructure, which then opens up new possibilities like dynamic HVAC or real-time availability of spaces or notify me when a place is quiet.

20:57or is that cafe currently super busy right now? Or we should like totally change the structure of our buildings because we've got 53 % of meeting rooms being used by a single person. All of those are outflows of making the infrastructure low cost and scalable. There's also an environmental impact here because you guys talk about on your website, knowing what's been used and therefore avoiding cleaning clean spaces. And so I can also see saving water, chemicals, paper, et cetera, in larger offices with lots of rooms and boats. Yeah. So about$6 a square foot. So assume like$30 a square foot is the average cost for a lease in the US.

21:35Premium Class A is about$60,$65 a square foot. Your facilities cost is about$6 of that 65. And so you can optimize that six with cleaning. Where the real returns are on energy or on, I'm sorry, on sort of sustainability is that 39 % of all global carbon emissions cost. come from buildings. Half are from energy, operational carbon. Half are from building materials in the construction of buildings. And we add to the net new buildings every year. I'm sorry, every day, 13 ,000 net new buildings every day. And that is going up. So if we just didn't turn on air conditioning and heat on floors that were unoccupied or areas that were unoccupied we could do a massive amount of not only um saving the environment which is awesome and virtuous we could also take load off the grid which we're having grid problems and then on top of that uh the person who runs the office and the facilities would save money it's like a win-win-win-win all right everybody you know my favorite browser that i use on my mobile phone and my desktop every day, all day, is Brave.

22:52Why do I use it? Well, it's privacy first. And I don't want people knowing what I'm searching for. I don't want to be tracked. I don't like ads. Really, it's for two reasons. One, I think it's creepy. But two, it slows me down. Loading web pages with pop-ups and ads, it's just insufferable. And Brave also has a built-in search engine, and it's awesome. Brave Search is powered by one of just three global-scale web indexes. And unlike other independent search engines, it isn't a skin over big tech technology. right? It's their own search. And they have 7 million daily users, 35 million queries, and it's the fastest growing independent search engine since Bing.

23:30And you don't need the Brave browser to use Brave Search. You can just go to search.brave.com using any browser. I mean, I of course think you should use Brave for so many reasons. It's also got a VPN built in, which is really cool, especially when you're at a cafe. So search all day long, every day at search.brave.com. And if you want to try Brave Search today, go to brave.com slash twist, get better results, protect your privacy, and get the answers you need faster. Go to brave.com slash twist so they know that we sent you and download the browser. It's awesome. What happens if New York City knew how it was used?

24:02That is kind of the core question. If you snapped your fingers and on Monday, New York City knew how it was used, had perfect data on use, would anything change everything would oh yeah like everything would change like you would change the subway pattern you would change how cars move you would change the cost of rent you would change where you would build buildings you would change how many starbucks baristas were at one cafe versus another just i mean and all of this is done with privacy so i think maybe you could speak to privacy here andrew and why that's important because this was something you and i had a bunch of debates about early on and i was like why don't we just use a camera for this So the answer to that question, yeah.

24:42Yeah. So this is, it's really small. This is the board. Ah, wow. And it's about 75 millimeters by about 20 millimeters. It'll be sort of like the final device. And you can see the renders and the final product shots on the website. If you want New York City or any city to know how it's used, you have to solve the distribution problem. How do you get an intelligent device into every relevant room in the world? One of the interesting barriers to that is if you're deploying surveillance tech, people get really uncomfortable using spaces in certain ways when they're observed. And the second that they feel like they're not being observed, and you see this play out in Airbnb, right?

25:26Airbnb has a global policy of no cameras allowed in any of their spaces if you are a host. And the reason they do that isn't because cameras aren't useful or because hosts don't want them. They do want them. It's because guests really don't, It doesn't drive a distribution of use on the guest side. And so if you can achieve the vast majority of what you need to achieve and use with an anonymous at source system, I think that's good for business. Yeah, I was in an Airbnb recently and they had cameras on the outside of it and they had it in the backyard where we were hanging out and there was a hot tub and a dinner place.

26:05And I'm like, that camera's on. And it's obviously could be recording. Every conversation I'm having is recorded at this Airbnb in the backyard. And I was like, how do I unplug it? And they had obviously strung the cable up to make it so you couldn't easily unplug it. And I was like, well, this has got to go in the rating system here. This is not cool. I mean, obviously, convert cameras. But yeah, it is very nice to know for a hotel if a guest is in the room so you can clean it. That's right. And this would be such a game changer. When you stay at an Amman hotel, they have a door monitor sensor system.

26:40So they do know a door opens and closes. If a door opens and closes, they will send the cleaning crew there to go clean it. But they don't know if you're actually in there or not. They just have a guess. Here, you would know. There's two or three people in the room, whatever. And I had an Airbnb for a while. And we had a no party rule. And somebody threw a party. I found out about it because I had a driveway cam. And in the morning I woke up and like from 3 a.m. till 7 a.m. cars and Ubers were dropping off. And I had like 30 notifications when I would normally have none during that time period.

27:13So it was very weird. Do you have the chart of the cost? There was a really cool chart you had with the cost. And I think for entrepreneurs, this is super important. You've been working on this company for a decade. and you know it's just another one of these uh as we call it in the business uh overnight successes 10 years in the making a decade in the making so maybe you could show that um so here we go cost per sensor over time and you had the depth sensor cameras or optical sensors radar the battery that's right yeah so so you can see the uh and you know a more detailed breakdown would include what portion or percentage goes to installation what portion is sort of core hardware software but ultimately the total cost of ownership has a huge impact on adoption.

27:59And what's interesting is like all, and you actually see this similar chart, you know, with SpaceX, they'll show a similar chart on like cost per kilogram to get to orbit or, or, you know, essentially a weight measurement of how much does it cost to get a certain amount of weight to orbit and all of the interesting stuff. I don't know if you can see my cursor, but all of the, all of the interesting stuff happens sort of in the bottom, right? um and so if you look at waffle relative to these costs um it just has like a really big impact on what types of new places you could deploy in the world yeah i mean if you're a cafe and you don't want you know a cafe owner doesn't want to spend this kind of money or a hotel you know doesn't want to spend the money this is a way to actually do that uh wow continued success here this is amazing andrew if you look back on your journey here as an entrepreneur um what have you learned you know, and you're, if you were to mentor folks who are starting a hardware startup specifically, because hardware is hard and your hardware and SAS and SAS is on the easier side of businesses.

29:00That's why a lot of people like to do them because you know exactly who you're going after. You get all that benefit. Um, and you have like a very narrow specific problem you get to solve and it doesn't have the, I don't know, um, randomness or lightning in a bottle of consumer, but man, adding hardware, hardware is hard. So as a hardware founder, what have you learned and what advice can you give other entrepreneurs listen? Well, in a world where, you know, you can ask Cursor to help build a piece of software and replicate, you know, split wise in about an hour and a half, I think physical systems, you know, going after hard problems and building physical systems is a hard problem.

29:40It will be defensible, much more defensible than I think sort of just like pure software layer. That said, but you have to really want to do the thing. We're wandering into a world where the infrastructure is so much better than it was when I started. There's so many new things that you can build with on really good rails. And the companies that are going to win, I think, are the ones that are willing to be playful. And you see this in Tesla. You know, you see this in Anderil. Anderil is a defense contractor and they're still playful. um and and i think the the folks that are willing to sort of like you know not take themselves too seriously but also build you know exceptional stuff are probably going to win also go to go to the office if you're a startup like don't do it distributed um and that's not that's not unpack that for me unpack that well what trust is built in person and it's lost virtually and and i have this theory um i was talking to someone just yesterday about this uh and they brought up a really interesting point and i i it never really occurred to me but you can't read body language as well distributed you can read some but it's actually deeply muted it's like wearing socks on your feet you know or gloves on your hands but as soon as you get in person all of a sudden like the conversation is funnier that you you want to know about their lives like it's sort of rounded out and i'm i'm reasonably convinced that it actually has to do with just like physical body language it's like somatic it's like somewhere deep in our human existence connecting in person uh resonates here's an image uh from the visual capitalist uh who does really nice graphics of um the cost of space flight and as you can see here uh the cost per kilogram of like the space shuttle delta heavy and then uh here comes falcon one falcon nine falcon heavy and then here's the starship estimate and what this chart shows is a slow curve coming down and then it falls off a cliff very beautifully so the slope just you know angles straight down this is the reverse of the hockey stick or it's an upside down hockey hockey stick cost plummeting And when costs plummet, adoption goes up.

31:57If you were to build a six-lane freeway and you previously had a two-lane avenue, what you'd find is people would induce traffic. People would get on that highway and explore new places. And that's what the highway system did in America. It just inspired people to go for rides because you could go 60 miles an hour and go somewhere 120 or 300 miles away from your house quite easily and comfortably. If you're a founder, like nail your unit economics and then do that. Because what happened in the run up, you know, in sort of the Zurp era is this would happen, right? But it was with negative unit economics.