In short
Podcast Notes: Relentless - Episode #31 with Garrett Scott, Co-Founder of Pipedream
Episode Overview In this episode, the host interviews Garrett Scott, the co-founder of Pipedream. They discuss the company's new innovations including their first Rapid Fulfillment Center (RFC), the autonomous delivery robot named Otter, and the upcoming introduction of Goods Grocery.
Key Concepts
Hyper Logistics
- Definition: Hyper logistics is described as the ability to deliver items in under 10 minutes at a minimal cost (less than a quarter).
- Impact on Commerce: This model allows for a bi-directional flow of goods, transforming how commerce operates by treating physical items similarly to data on the internet.
- Potential Benefits: The implementation of hyper logistics is positioned as a significant contributor to GDP growth and overall value creation.
Evolution of Pipedream
- Initial Vision: After taking time off post-college, Garrett aimed to address a significant problem in logistics that he was passionate about.
- First Iteration: Originally focused on drone technology for logistics, the vision evolved to encompass a broader understanding of hyper logistics.
- Market Gap: There is a perceived lack of competition in the autonomous logistics space, with few startups addressing the same challenges.
Delivery Experience
- Changing Consumer Behavior: With faster delivery options, consumer behavior is expected to shift towards higher frequency, impulse purchasing (e.g., ordering a single toothbrush instead of bulk buying).
- Retail Experience: The idea of "teleportation" of goods leads to quicker decision-making and a more efficient consumption model.
Innovations and Developments Rapid Fulfillment Center (RFC)
- Purpose: The RFC is designed to enable rapid order fulfillment and decrease delivery times.
- Location and Design: The RFC is strategically placed in Austin, designed for high traffic and accessibility.
- Operations: Orders are expected to be filled and dispatched in under three minutes.
Otter - The Delivery Robot
- Functionality: Otter is designed to navigate an underground network to deliver goods quickly to surface-level kiosks (portals).
- Operation Speed: Capable of moving at speeds up to 100 miles per hour.
Goods Grocery
- Launch: A grocery service launching in conjunction with the RFC.
- Stock Strategy: Good's Grocery will offer a selection of high-demand items that can be picked up quickly, addressing the issue of key items that customers might need.
Challenges and Considerations
- Logistics Complexity: Building a new logistics network comes with inherent challenges, including navigating commercial real estate and regulatory hurdles.
- Customer Experience: Creating a seamless customer experience by reducing friction in the ordering and fulfillment process is paramount.
- Talent Acquisition: The importance of hiring top talent who align with the company’s values and vision; Garrett shares insights on the hiring process and maintaining a high talent density.
Company Culture and Values
- Team Dynamics: Emphasis on a team culture that prioritizes high-caliber individuals who care deeply about the work.
- Customer-Centric Philosophy: A commitment to creating a great customer experience, even in the face of challenges.
- Personal Stories: Garrett shares anecdotes about team members, illustrating the company’s emphasis on empathy and support within the workplace.
Future Prospects
- Scalability: Garrett discusses the plans for expanding the network of RFCs and their potential for rapid growth.
- Market Position: Pipedream aims to be at the forefront of the autonomous logistics revolution, setting standards for speed and efficiency in fulfillment.
Conclusion Garrett Scott's insights into the future of logistics illuminate the exciting developments at Pipedream. The focus on hyper logistics, innovative delivery methods, and a strong company culture positions Pipedream for success in the rapidly evolving retail landscape.
Written by AI. May contain mistakes. Listen to the episode to check what was said.
Transcript
Automatic transcript. May contain errors.0:00I like to solve one very hard problem for a very long time. Getting things to you quickly is just something I've always loved. Like, I got goosebumps the first time I used Postmates. I just couldn't believe it. It's just someone brought me coffee. The amount of bandwidth reduction on doing things is just so high where you can just, like, hit a button and something happens. Does this solve the kind of problem where, why isn't there an Uber for screwdrivers? It's too slow and too expensive per delivery. We define hyperlogistics as the ability to get something delivered in under 10 minutes for a de minimis cost.
0:26We usually say, like, less than a quarter. And once you achieve that, then there's this huge explosion of value creation that's possible. Garrett, it is awesome to talk to you again. So you're one of the co-founders of Pipe Dream. And I think over the past like three-ish years, something like that, you've gone through a whole bunch of different iterations on what you call like enabling hyper logistics. Can you explain what hyper logistics actually is? Yeah. So we define hyper logistics as the ability to get something delivered in under 10 minutes. for a de minimis cost. We usually say like less than a quarter, but just like so low that you don't think about it.
1:04And then the ability to as easily as you receive from that system, you can send back. And what that enables is bi-directional, super fast, basically teleportation of things to you, teleportation of things back. And what we think is it unlocks a new level of commerce where you start to treat objects like you do data on the internet, where you get things delivered when you need them, use them, and when you're done with them, you send them back. Yeah, we think it's going to unleash all sorts of um i i yeah in the long term i think it uh is the number one increase in gdp um from this episode that's that's what we're working towards how'd you like come up with that vision because i imagine you know there's a bunch of different companies that have just kind of like tried to enable you know a better delightful uh customer experience with buying lots of stuff um what was the initial like idea that you thought of and uh what's been the kind of transition from the first iteration of like the drone idea to what you have today yeah so after college I took
2:05took two years off I did some other stuff but long story short I had two years where I was like you know what I learned a lot about going zero to one and I learned enough about myself to know that I like to solve one very hard problem for a very long time that's what makes me just innately happy and so I decided you know what I need to take some serious time and decide what that problem is the thing that is at the intersection of a problem that is worth solving for a long time and has a long time of solve in it and something that I wake up every day and if I fail for 20 years I'm just so happy I was working on that problem because I care about that so much and so looked at a lot of things during those two years and getting things to you quickly it's just something i've always loved like i remember the first time i got goosebumps the first time i used postmates and like i just couldn't believe it just someone brought me coffee and it was the sickest thing i was like oh my gosh when was that uh
3:022015 maybe it was in oklahoma city so i think postmates took a while to get there yeah the first time i used it's like oh my gosh this is amazing i remember the first time i used an uber that was magical uh i berated the guy um i sat in the front seat and i just berated the guy Like, okay, how much money you make in? How's this work? What are the problems? Like, just the fact that I can hit a button and a car comes and takes me is like, oh, this is insane. The amount of bandwidth reduction on doing things is just so high when you can just like hit a button and something happens. I remember the first time I used Amazon, I was at a friend's house and you ordered something.
3:37I was like, that's going to be here next week. Like this whole library of things is just going to like, you can send these things to you. That's crazy. um and i just like i knew that's something that i loved um it's also uh it's one of those things that we forget about it but like every time you make delivery things faster it impacts society in such positive ways and it changes our lives in so many ways and we forget about it yeah um if someone tomorrow took away next day amazon it went back to two days or like even three days like the amount that impacts my life is huge the amount that impacts my startup is huge the amount we rely on Amazon being uh able to fulfill our orders by next day is crazy like there's so many shortcuts we can take because we know we have reliable access by the next day yeah totally um and so that that's what I realized like you know what that is at the intersection of I think there is so much progress left to be had there we've been um over the last like 20 years There's been only kind of incremental improvements.
4:40Like even Amazon is kind of very, very hard what they do. But it's really kind of an incremental improvement. I would say like Uber Eats and DoorDash. Not quite a step change. A little bit. Those are also incremental. And even those had huge impacts. And it's just like, man, we have so much room left between Amazon and Uber Eats and where the future will be. That's what I want to focus on. I was like, okay, like clearly autonomous logistics will be here at some point, right? And to me, it was like, that's the bet that I want to take. Like autonomous logistics, it is for the impact it's going to have, there are so few people working on it.
5:22And to this day, I can't believe it, how few people are working on autonomous logistics relative to the impact it's going to have. And in that exploration, because I had two years to think about it, we realized that like once you achieve autonomous logistics there's this new stage of logistics that comes right after that that is quickly enabled there was a name for it we named it hyper logistics and once you achieve that then there's this huge explosion of value creation that's possible do you expect that like people's kind of frequency of just ordering things on the spot is just going to go way up and so instead of having a cart with 10 items you just buy a toothbrush and it shows up in 15 minutes kind of thing
6:07yeah maybe not a toothbrush but yeah i mean you probably don't um you probably don't keep toothbrushes like right now we probably you know i have i have backup toothbrushes i buy them in big packs probably don't do that as much um i think there's uh i think you you can go for a higher quality toothbrush um just get them as you need them um but i think where it really impacts are things where you go to order on Amazon and you pause, or you go to order on Uber Eats and you pause. Those things stop becoming pauses. You execute them immediately. For instance, like close buying on the internet. Instead of really rigorously looking at all the things that you might want to buy and thinking about, well, would that fit me?
6:51Would that look good? You just order everyone that looks good. You can send them all to you, right? Because it's so cheap to get them to you. try them on in your house and then you know 10 minutes later you know pack them back up and pack them back and send back just throw them yeah and just like yeah keep the one that you want and then after you wear it that night uh you can keep it or what i think is like you know you'll create these really efficient um ecosystems where you're like if i ever want to wear that again i'll just have it delivered again send it back right um and clothes become a subscription and you never wear anything twice and like me and you are wearing like super high quality shirts because it's just like we're not paying full value for these um i and that's i mean that's true of content right like uh the amount of access that you had even like something as small songs like someone who is really rich um their ipod had orders of magnitude more songs than someone um who was poor right their ipod would have you know i remember this i had 15 songs on my ipod and i was like saving for the songs man and i would just i just go through them i just had 15 songs i think i i got a uh uh i got it off amazon or maybe it was ebay i think it was ebay i got like a uh remember those the ipod that would clip on you yeah i got a knockoff one of those that was basically just like a flash drive and just had like a play pause button i got one of those and i had 15 songs and i was like feeling good right um and then like you think about today and everyone has access to all the same songs with spotify with spotify yeah yeah and i think the same thing happens um with objects it's a little harder to think about um because we think about objects as having like this uh finite value right like the distribution cost of songs with the internet is way higher but i don't think the delta is that big does this solve the kind of problem where it's like why isn't there an uber for screwdrivers because it's just too complex and it's like doesn't makes it's too slow and too expensive exactly delivery yeah exactly um because you're you know you're looking at like 12 to get that screwdriver delivered so that's that's a huge delta on top of the screwdriver then if you want to send it back it's another 12 so you're 24 in and it's like i could have had a really nice screwdriver for 24 yeah yeah totally it just doesn't make any sense um it barely makes sense for food um barely but even that is huge like people pay a huge premium on food to have that delivered because it is so convenient so once we drop that um drop that cost i mean for food it's gonna go crazy yeah uh just fresh food every day kind of thing yeah fresh food every day yeah totally whatever you need when you need it um you make the decision where you're uh gonna eat in the moment um your pantry becomes what's available on the internet like you don't have a pantry right you're real timing chips to everything i think so um and that's not true of everything like you'll want to buy eggs in the dozen probably because it's nice to have an egg available instantaneously like that to men delta means something um i think there's like items of um obviously like sentimental significance that you want to have i think there's things that are too personal i'm not going to like rent underwear uh but you know maybe i i get those items um laundry service starts to make a lot more sense when sending it to the laundromat or what I think will be, I think there'll be like a cloud hub for those kinds of services where you send your clothes in a bag, they get washed, they send back to you folded because the cost of logistics is the only thing holding that business back.
10:26So yeah, like TVs, electronics, like those kinds of things, you're not going to rent. But I think the amount of things that you do rent just goes drastically up. drastically up because the access to quality how much more quality those things are when you can rent them when you need them versus if you buy them at full price is so much so much higher interesting okay so we are sitting and i think you're like first distribution center right now it's just a warehouse and i think you got this like a week ago or two yeah um but you know how what's it like you know getting this first place and what's going to happen i mean getting getting someone to Lisa's commercial real estate was it's always so tough for startups and it's always every time you pull it off It's like a little miracle Do you go to other leases like they just didn't pan out at all?
11:13What was that like?
11:17Well, we like we need what we wanted was a fulfillment center of significant size That we could turn into what we call an RFC, a rapid fulfillment center, right? We want anything stored in here to be able to be parallel picked, that order combined, and out the door in under three minutes. And so there's some things that we need special. But we also want to put the first portals in the parking lot. So we wanted something where this location is north of downtown in Austin. There's 30 ,000 people who go by every day, right outside. So like finding somewhere that is well trafficked and good demographic because you know this is difficult i imagine these portals are basically going to be like our first store right so you wanted to you want a fulfillment center in the city uh that is close to the right demographic they don't build fulfillment centers uh so it was tough there were very few spaces that actually worked and then of the ones um that did work uh fulfillment centers are zoned for i mean you you know you can build like uh uh you know know 10-story hotels on top of the film center so the land is so valuable so like especially in awesome where it's booming like people don't want to lock up a space for five years uh and then if they are going to lock it up for five years they do not want to give it to a startup uh especially one doing something that you know uh in their minds is way too crazy never gonna work um and so it's super hard so tough uh and it's not like it's not even quite like a distribution center because a distribution center you can kind of just throw wherever because you have the trucks arriving but you're actually going to have people picking up their groceries i believe from the portals yeah outside so they kind of needs to be in the proximity it needs to be in a really good retail location so you need a fulfillment center in a retail location right um and also like this is going to be the um where we build the network out from so it needed to be within where we wanted to build the network so i in the future um we're going to build the rfcs outside the city and then come in um but for this one we wanted to make sure that we were able to put the portals the first portals in the parking lot uh really quickly um and wanted to be in the good location so it was like the the parameters we were working with were um tight yeah but we did we found like multiple places that worked and then it was just uh the discussions people to let us do it so we did end up i think i think we ended up with everyone that we wanted eventually did come around to letting us do it but uh was it like a month plus of negotiation and talking with people it's probably like three weeks three weeks yeah yeah um but i mean canon's so good at it my co-founder um of talking to people and um you know i another tricky thing about this spot in particular is uh we are taking up a lot of the parking lot um for our portals and kind of like all the traffic needs you need for the portals we're putting three portals so it's a pretty big um we're building this beautiful canopy over them as well um because we really want this to be like a show piece um which means we are taking parking lot spaces from uh our neighbors who are used to parking spaces so um we you know we had to convince them and um they're great like we got so lucky with our neighbors but um yeah a lot of uh a lot of things to make this work a lot of things had to happen on extremely tight timelines uh exactly right had to be super well coordinated uh we have the scan chart that's just like huge and all these things had to happen at exactly the right sequence they were all blocked by each other um and i was just like our team can just go demon mode when we need to and like we have to have this thing so let's like it's not only elon it's actually the entire team going demon mode totally yeah yeah every individual person can you know do their own version of demon mode um do you want do you want to explain kind of what like an otter is what what this distribution center will look like and then what an otter what a portal uh and then also instant pickup just so people know what those are totally um so pipe dream what we made is a system that can um get things from point to point in a city uh really really quickly so um we have the robots we call them otters um they can carry uh it is a tote about yay big uh it's the walmart standard tote that's like so what walmart uses for their e-commerce that's what we're using for um this and the otter holds it um and can go through uh our underground network at 100 miles an hour so we build these pipes that go underground um and 100 miles an hour they can bring things to our above ground kiosk that we call portals and deploy the payload there.
16:04And these are just like places where you can pull up with a car and it's a little bit like a bank, you know, has the money. Yeah, totally. Disposable money. And you just like roll down your window, throw the groceries in your car and drive away. Exactly. So if an ATM is a tiny bank, yeah, right. A portal is a tiny store. So that's what we think about it. A portal, you can do all the functions of what a store is a modern store is now um stores have become uh warehouses that are in the city and well positioned right um and you get things delivered from those stores you can go and pick up stuff at those stores and then you can also shop in those stores uh the problem with those stores is that um using them as a warehouse uh naturally reduces the experience of the in-person shoppers right if you really do it right yeah so no one wants to make that consolidation even though pickup and delivery are their fastest growing verticals.
16:56So they are still making consolidations for the in-person shopper. I think in the future, it's very clear that pickup and delivery are going to be the only two methods that we shop with. So our thought was really the future of stores should be focused on how do you make pickup and delivery as fast and as good as possible. So that's what the portal is designed for. it's either handoff to a delivery driver handoff to a person handoff to a drone handoff to a sidewalk robot um they are both stores you can come pick up from if you're a human being but they're also a distribution node um so they are essentially a micro performance center um is it also is part of the thinking you know eventually if you do integrate with walmart or something there's no it's just like hot swapping your technology into their system because it's the same tote kind of thing is that oh yeah yeah for sure it's a very standard tote so there's a lot of retailers that use that tote.
17:49So that's how we landed on which tote to use. Amazon is an outlier, but their tote fits in the system, but it's like a little different dimension. Slightly smaller. Everyone's using about the same tote, because once you reach that tote size, I think it's 99.6 % of items fit within that tote by volume. So it's kind of a suspiciously large tote. You can fit a lot of things in there. So like a 60-pound dog food bag fits. 48 pack of water. Yeah. 24 pack of toilet paper. Fits in the tote. How did the evolution of like the Pipe Dream idea from getting, you know, minimizing the amount of time to actually shipping products all over the place to people?
18:34How did you kind of think through the process at the beginning of what you were initially ideating on and then have that change over time? Yeah. Yeah, what we are doing now is outside of one difference, which you can get into, this RFC and then sending things to the portals is what was in our very, very first pitch deck. It's a cold start problem. It's a huge cold start problem. It's like massive. Like, how do you cold start that? So the first thing we did is like, we just need to get something in the ground. So for year one, we permitted an in-city system in Georgia. And we said we just need to like, everyone thinks we cannot do it.
19:20And that was like a Wendy's partnership, right? No, no, no. This was just with the city. So it was like a mile long system in Georgia. We started construction on it before we had finished the first version of the tech. And so we were making a lot of tech decisions on the fly. but permanent in two weeks, built the system. It was going from inside a retail store to inside an office building and taught us a lot about what matters in the system, what does throughput look like, how fast can you really get these things to go, how easy does this construct? Because we had confidence based on the heuristics of underground pipe is not difficult to put in, but it's like we've also never done it so we you know we're just going off the word of other people so um give us a ton of confidence in underground construction um it gave us a lot of confidence in permitting because we were able to permit that one in two weeks um was that like a false positive did did the other permitting end up being easier or harder no permitting has always been it's been roughly pretty easy yeah and is that because you're bootstrap or like uh hopping on top of the like whole sewage you know yeah it's just like for underground infrastructure there are known pathways to permitting.
20:37It's not the same as like drone delivery where it's just like extremely difficult to get the air permits and stuff. Yeah, totally. It's like if you were doing drone delivery eight, nine years from now is what underground permitting is now. Right now. Where it's like there's two process, but it's not much because it's underground infrastructure because the nice thing about... What's that? It's like overcomable. I wouldn't say it's overcomable. It's a process. Okay. The process to put in underground pipes is just a process that everyone's aware of. There are decades of stipulations that every city's put in.
21:07There are well-worn pathways. So doing it is a known process. Whereas with drones, that's a new technology. Pipes are not a new technology. So drones, like, you know, right now we're all building kind of what does that regulation look like? And then once we land on that, then there's, you know, due process. The difference a little bit is, and this is, I'm the biggest drone fan in the world. They're going to be massive. people are underestimating their impact to like to this day underestimating by maybe one to two orders magnitude yeah cannot believe that people are still underestimating them I think it's one of those things where people just thought that it would kind of be the obvious future and it just hasn't arrived yet and so they're kind of like oh well is it actually the future if it's you know we don't see it every day yeah totally I mean it's that adoption curve they're in the Valley of what's it called Valley of disillusionment or something Yeah, yeah, yeah.
22:04But, I mean, yeah, that disillusionment is creating a lot of opportunity that I can't believe no one's tackling right now. Neither here nor there. We should get back to that, though. Yeah. But, yeah, so it's totally different. The amount of impact that underground work has on a city, everyone knows, is minimized, especially because we tell cities we're only doing horizontal directional drill. We won't trench. So you probably saw some of these machines. there's some on the road right now they just sit on the sidewalk and they dig underground and they pop up like yeah it's like a thousand feet to like yeah what's that it's like the big you know boring Co it's like that but the smaller yeah you start with this and then you grab what is basically and it increases the the bore size as you you go with something that's like this big rod this big it goes underground comes up you grab uh that hole extender it's called a reamer um and then you pull it through and that creates your your hole and then you push it through one more time you grab the pipe and you pull it through um and so cities know that's not going to cause like problems you can do like 24 7 they don't care um and there's no like once it's underground there's really nothing that can it can't hurt anybody so these are like really low stakes um permits to approve so I think that's probably the biggest misnomer people have it's just like and like as many times and we've done it like six or seven times now it's actually not that complicated no it's never never taken more than like three four weeks okay awesome yeah how do you think about from like going from an empty warehouse to actually turning on you know the first distribution center what does that look like yeah I gonna take uh so we're gonna the biggest thing The biggest problem has been finding our anchor tenant.
23:54Because every network, it needs kind of that killer app. And so we've been working with retailers on the instant pickup side. So instant pickup is the system that we put in that is not part of the network. It just makes your pickup faster on your own location. So we have one in a stick and shake in Arlington. We have some more that we're putting in.
24:20and through that we've been able to work with pretty much every big retailer we wanted to which has been super nice and so we've gotten a lot of face time with them and our original thesis was you know let's work with one of them let's not try to do too much like we want to be base level infrastructure let's work with one of them to be that anchor tenant from working with them um we realized that you know i i don't think on a network like this anyone's going to care about as much about the customer uh as you do yeah yeah and so um we realized that like these bigger incumbents are not set up to be the anchor tenant on network um what it really takes is someone who is so laser focused on this type of customer experience and creates a business specifically focused on taking advantages, taking advantage of like what this network gives you and just really tightly iterating around what these, what autonomous logistics can give the customer and what the customer wants from autonomous logistics and just like really tightly focusing on that feedback loop.
25:31um and so this is something that we kind of realized right when we started but it was just like don't don't want to so let's try not to let's go you know try to work it's like a more painful path right if you're not plugging into it's a more painful path and i think when we started we considered it to be a more risky path um and then over the last uh like year and a half we realized that we actually think it's the less risky path that that having full ownership over that feedback loop is where all the value is going to get created. And so, like, ultimately, we do want to create the network. That is what we're focused on.
26:09We, you know, I, to this day, I think the vast majority of Pipetune's value over the next 10, 20 years is going to come from fulfillment, right? Operating like a 3PL, people plugging into the network, delivering on it, like all that volume is where most of the values can be created. But I think to kickstart that, like doing that first party retail ourselves is what is going to grow the network the fastest. Really, really, I think it was maybe like six or seven months ago. The real kicker came when we decided like we absolutely have to do that is when we started to run the numbers on per portal.
26:51If we're doing first-party retail, we're running our own retailer, how many orders a day per portal do we need to go ROI positive on the network? And we realize it's 42 orders. Oh, that's not a lot at all. No, and that's to cover... How many can one portal do per hour? Like how many swaps? It only takes like a few minutes, right? We've got it down to like someone can drive up, grab it, and go. um and so depending on how it really depends on how much someone lingers that's the choke point not anything on the tech side like we've got that thing snappy um so after you get your groceries it'll just like pop up a little screen saying all right now you need to leave something we'll see right right now we haven't found that people linger that much uh if we need to we will though um but i mean people are like you know you're using that to go quickly right it's kind of like people don't linger at a drive-thru you have like this social pressure if there's someone behind you you're like okay i don't like linger here i'll go and do my thing over there um so i don't think people linger as much as you think they do um but i mean like you probably at like probably the fastest you ever go is like 150 um it's like a 15 second interaction then you assume there's like 15 seconds on both sides for people linger but uh we want to get that as close to that 15 second interaction as possible so there's some some uh i guess we can talk about it um so on the portals out here uh when someone comes for the first time we offer them a um right now you have to use your phone to like say that you're here right it's such annoying friction yeah anytime you have to like pull out your phone to like say i'm here um super annoying um and there's three lanes we We've got to know which lane you're in.
28:35And so the first time someone comes, we're offering them a small RFD tag that they can put on their windshield. And so that we can see that as soon as they enter the property, we know that they're here. We get their order out to there so that by the time that they pull into that portal, it is out by the portals. And as soon as they pull into a lane, in the next 20 seconds, we've got the first one coming through. And they don't have to touch anything. They just get out of their car, grab their groceries, and pull away. Crazy. So we're going to get as much friction and non-joy out of that experience as possible.
29:09And then at the same time, one of the things that we are most focused on is because we're creating this kind of new store, that there's not really a human touchpoint. There's no human in the loop. How do you inject more humanity into the experience? More like delight. More delight, yeah. How do you make it feel more craft and more human? And like you can go to a museum and if that museum was empty and there was no people in there, it still feels like a human experience. It's just like there's so much intention and like humanity in there that is like, you know, reflecting back at you. And so I think it is if you're going to build an automated system, you need to build as much humanity back into that system as possible.
29:53So are you almost trying to think like, I mean, I don't know if this is the right way to look at it, but like almost a less futuristic thing. because I imagine that the more futuristic he is like less human in the loop it's just a robot doing it you know yeah yeah I mean we we debate we debate the um yeah and I think canon's right here but me and canon debate about how much of the story of the interaction is like the robot in the tech yeah and you know I I think I may be a little too aggressively the wrong way I like take it all out i just want that thing to appear yeah who cares how it happened right and i want to just be so joyous you just want it to be kind of like a magical experience totally totally yeah um i think like canon's more right like i think the how it happens is an important thing like if you were at um like a bank tube's a good example of like it's kind of like a magical experience like when the bank tube comes and your little like capsule comes out yeah um that's so fun uh if you made all that pipe opaque it wouldn't be as fun no even if you knew how it happened it just wouldn't be as fun so i think like there's a blend of how much do you expose how it happens to the customer and how much um you know you shouldn't rely on that i think that's the key is you don't want to rely on not everyone is going to think a robot's cool nor should they right um really like anytime you're using a robot somewhere should just create a better experience outside of its novelty.
31:19Anytime you're relying on novelty, you've under... I imagine that the early adopters will enjoy your robot. I hope that we have built a service where every early adopter hates the robot. Really? 100%. Why? Because that means that we are giving so much value to people that they're using it despite the automation. But we're just giving them so much value. right if if if 80 of the people are there love the robot and the first thing they tell people is oh it's like a robot comes out to you then we haven't built the great experience yeah um the first thing people should tell us is like i don't even think i just i drive up i grab my groceries i go it's like you know we're 10 minutes faster than any other grocery experience like uh i don't know how many people pick up their groceries but it's it's a huge industry um Um, and really every single other experience has a delivery window.
32:14You have to say what time you're going to go and pick up your groceries. Um, and then if you miss that window, you got to reset another window. It's like super annoying. You're like, kind of like you're scheduling yourself. So is this going to be something where you basically like pre-pack all these totes? And then as soon as someone arrives, their RFID tag says you're in the parking lot. Sends it out. And it just sends it out to, and then maybe it like tells that person which lane to go in or something. No, no, no. Or pick whatever lane you want. Really? Yeah. Yeah. We want you to be as mindless as possible, but you pull in, pull out.
32:44Kind of similar to a gas station. Yeah, yeah, yeah. Exactly. Or a drive-thru. A drive-thru is a pretty mindless experience. We just want it to be effortless. And we'll just keep pulling stuff out until it's as effortless as possible. So yeah, once you put it in your order, just come whenever you want, and we'll have it ready. That's awesome. How do you think through which products to actually stock here versus not? So when we decided to do this, we ran like a three-week test where when we committed to it, we built a grocery delivery service in, I think we got it up in like two and a half days.
33:20Really? We built a website. We scraped. When was this? I don't know, a couple months ago. A couple months ago. You just randomly decided to do this? I don't know. I'm so bad with time. No, I just like, you know, we had to touch the customer. Yeah. Right? Like we needed to learn things ASAP. Um, was that the most efficient way to learn things? I don't know. Maybe not. Um, well, that's one of the things that I really enjoy about you is that you, like, I remember from our last interview where you just talked about, I don't know how many different projects you just randomly decided, okay, I'm going to start doing this.
33:52And then you just started doing it. And, you know, maybe you decided to stop working on it at some point because it just wasn't worth working on, but you just actually took action. Whereas most people don't take action. Yeah. Yeah. I mean, we're, that's always a good thing. but i think i think like if someone is biased for action and someone is calculated nine times out of ten i'm gonna bet on the person who's biased for action right their likelihood to hit things are much higher i think like the best performers in the world are a bit more calculated like their hit rates a lot higher um but under those like top one point zero zero one percent there's a bunch of people who are just biased for action and then you're back to the calculated i think like Like there's a bell curve of, you know, at the very, very, very far end, there's insanely calculated people.
34:39But even then, like they're, they're very biased for action. They're very quick to admit that something's not gone wrong. I think like in testing the grocery delivery, like we built the site, we got in the hands of a bunch of people. We went and handed out flyers, like we were offering 15 minute delivery and we were offering it for free. We just wanted to like create a lot of orders. Yeah. We had scraped 7-Eleven and CVS's inventory. And that's what we were offering. We were like drive this 7-Eleven, grab the orders and deliver it to people. And this was a few months ago? Yeah. What? Yeah. How many orders did you get?
35:15We got one. You got one order? Who was the order from? You're like, mom? No, no, no, no. It was just some random guy. It was amazing. Yeah, yeah, yeah. And I think that's where it was like, it wasn't. How many flyers did you hand out? I don't know. Hundreds of lunch. Like a stupid amount. But we learned so much from it. That one order, we learned nothing on the one order. I don't know why that guy ordered. But ultimately, I do know why that guy ordered. Because we had the thing he wanted. And what we realized is when we were watching people go and search and look for their thing, there was just like, if you didn't have one item, you were shot in the foot.
35:57This is the same thing that I've seen with like Sam Walton and Walmart. It's something like, I think Walmart has some crazy number of like 130 plus thousand, you know, different items in the store. And all like 4%, there's a whole bunch of items where they have to have them because, you know, one person will go to that store instead of the other store because it has that item. Whereas, but no one else buys that item. Yeah. But it's worth having just so that it can get the extra few hundred dollars from that one customer. Yes. And so having, you know, ironically, having like only 20, 30 ,000, you know, different items in a store may make it very difficult to get people in.
36:34So how would you solve that? How would you make sure you 100 % solve that? And we kind of arrived on this idea through this process. but what we did is like um we we as soon as we have an idea we go and i have this big group of people um that i've kind of curated over time that i know give me exact what's in their head answers as soon as i say something so i was like i say something tell me exactly what you think like right now they they're just like shit testers kind of thing yeah it's more like um they are very connected to their immediate emotional reactions to things um so there's no like uh like the worst people to get feedback on an idea from sometimes is other people in the business uh because they're aware of so many variables yeah right when you when you when you mention an idea if we work together and you mentioned like you know a podcasting idea i would immediately the first thought is how do i implement this and then is it a good idea like i just can't get that out of my head right and so it's like you're so biased yeah um but anyways I have these people that I go to that is like you take ideas to and I know they can give me like immediate emotional reactions of what they're feeling.
37:44Yeah. And like useful feedback, useful feedback. And this was this this the way that we ended up solving this problem tested crazy well at demographic. And Cannon's got a similar process and like it tests really well with him. And everyone we've tested it with is like, OK, I think there's something here, something we can uniquely provide and make better over time. So how would you solve the problem? There are key items, especially in grocery, that if you don't have, you will lose everyone's business. However, what that has also created is a fragmented system where people have a few of those items at a few different places.
38:21So the amount of stores that people are going to is a lot higher than you would think because especially families are going to multiple different stores because there's a few of those things at each place. So we're starting a new grocery concept. you know we're taking we scraped every single grocery store um all their products and we created this subset of the um i think it's 42 000 of the most uh that that the um the most stocked items items yeah yeah that's in the middle these are like oreos and um doritos doritos coca-cola yeah and like the little packages of meat like we've got that subset and that's what we're gonna I'd imagine also like fresh, you know, groceries and stuff.
Read the full transcript
39:05Yeah, yeah, yeah. Fresh vegetables. They're staples. So there's like common things that, you know, you just have to have. So just talking all those here. Those are available within five minutes of ordering. So as soon as you order, you can come pick it up in the next five minutes or take three days and come pick it up in three days, whenever. And you just basically like, will there be like a Pipe Dream app where you just say, okay, I'm, you know, showing up in five minutes and here's my grocery order. Well, Pipe Dream would be a horrible name for a grocery store. So we went through hundreds of names and landed on goods.
39:35So goods is the brand. It does have kind of like this retro feel. People have been loving the brand. I'm just going to go to goods. Yeah. That's where you get your goods. Yeah. Is that goods? Exactly. We have good goods. And eventually, we know we'll do all goods and we won't do services. That's where we draw the line. That's going to be another Pipe Dreams spinoff. No, no, no, no. We're not touching services. Only goods. Only goods. Yeah, so how would you solve the problem? We have the top 42 ,000 are here, stored here. Though there are key items that are in these other stores that people care about.
40:14How would you fix that? Honestly, I have no idea. I imagine that there's probably a lot of items at Walmart that you can't even stock here because it doesn't really fit in a tote. And so you probably just immediately take out all of those. Not in the grocery category. Okay, not in the grocery. I mean, okay, so Walmart has a bunch of other items, though. So, like, how many is it typically in a, you know, grocery, you know, grocery part of Walmart? Oh, how many shoes are they holding in the grocery? 130 ,000. I think it's in the 60s. 60 ,000. Okay, so you're able to hit most of it. Canon would know, I don't know.
40:47Like, two-thirds, right? Ish. Yeah, I think so. Okay. I don't, I think you'd probably want to take a more convenience store approach. um because you know you already know that people will show up without you know just for a few items select items yeah um but i could be wrong yeah i how'd you think about it that there's one way we thought about it going on the um there's a there's a there's a blend of grocery it's like convenience grocery um it's maybe like where there's stores that kind of sit in between like like they're a corner mart right they have like a good allotment of things and some like cool things A lot of startups have gone that way.
41:29I think the parallels would be like DoorDash's Dash Mart has done really good. GoPuff would be that. GoPuff has expanded. They're maybe closer to actual grocery now. They've expanded their SKUs. The problem when you sit there is you kind of end up taking that 30. It's 30-70. So 70 % of like someone's grocery shop happens on like very prescriptive grocery shopping. Yeah. And then when they need like filler items, that sits in the 30%. So convenience grocery fills that 30 % void, 30-ish percent. I think there's just so much danger in there because you can't really build a habit. Yeah. You spend a lot of money in attention to make sure you're on their mind when they need something really quickly.
42:20but it doesn't necessarily build brand loyalty. And you see that in like, I mean, like even internationally, the companies that are doing quick commerce where they're just focused on that core set, it's really tough. You have to constantly win that customer over and over and over again. And it's really expensive. Now, if you look at, and then that's where initially we're without going. That's like a well-worn track, like it works, right? And it's easier to do that. What we really wanted to push ourselves is how do we capture someone's entire grocery spend? We want to be that 70 % and then earn the 30%.
42:53Because if you can capture that 70%, you're winning that customer. You are winning that customer and the ability for you to win them in other categories goes way up. When you're convenience, people see you as convenience. They see you as lower quality. They see you as putting a premium on top of the dollar because you're creating convenience. What we want people to see is we are not putting a premium. We are keeping those costs low and you can trust us with that full grocery shop. So we can win you there. Then we can, what we want people to see the portal as is kind of like an everything store.
43:27Yeah. Right? I can go get a lot of things there, right? And that builds kind of the network effects as other people come onto the network, as we stock other people's wares in here and we do 3PL for them, people see that portal as something you can go and there's a lot of options but um are you gonna try to do like ready to eat meals as well because i'd imagine i'd imagine that like if people you know if you're buying from amazon it's not quite the same but if you're like showing up the costco hot dog is a massive driver of why people go to costco same with like the rotisserie chicken totally um that's that's what we call the special item every good retail play has like this weird special item so is it going to be like kirkland signature but it'll just be like special goods you know signature version yeah yeah um yeah i we're locking uh we have some really good ideas for the special item i think we'll wait until we open before we talk about them but like there's some really good ones um that i'm excited about you have to have a special item they're super important so ready to eat meals um we're not going to roll out with um but we have a really good plan to do in q4 okay um it's a little different than you would think but um wait so are you saying like in q3 of 2025 people will be able to pull up to a portal here yeah and get a grocery order so the gantt chart has us at 55 days we have a clock in the office that is like 55 days till till we open so like there are things that could go wrong i could told you we've been so tight on that schedule so tight on it um something good could go wrong yeah um from here but the like the things that really could hurt us are behind us um so like the big uh milestones that like probably like when we laid out the plan we're like do you want to talk about in there that we cannot do do you want to talk about a few of the things where permitting yeah permitting getting commercial lease um we built this um we're building this like gorgeous canopy uh that is going to sit out there over the portals and it's it's gorgeous and uh we didn't think we could afford it um and so we ended up just calling like so many canopy companies to find someone who could build it for us for like a cost that made sense and we finally did and i like could to this day cannot believe we're getting it at the cost that we're getting it at um but we just like what's it made out of um it's made out of just like steel and then uh there's this type of um starbucks uses it a lot uh and we actually the tennessee Tynas Stadium almost took the entire supply for it and so we had to like get into the deal.
46:04Hayden who does he's had a product for goods and he just had to call so many people he wanted this type of it's aluminum so it's like really weather resistant but it looks exactly like what it's gorgeous what Starbucks uses for their wood elements a lot of the times and yeah the Tennessee Tynas Stadium was taking like so much of the supply so we had to like really fight for a supplier and there are a lot of things that like we're not going to like half ass something you know we wanted to if you're going to do it do it really well full and a half ass something right so the lights that we were using to the temperature to like the way it laid out like we didn't want to make any consolidations because we wanted to do it quickly we wanted to both do it quickly and do it really well so there's some of those that was just like the timeline seems really tight to do that really well but everyone like just cranked uh Peyton who's head of product is so good at paying attention to details and making quick on the fly decisions and like the brand he created and the feeling he's created uh around the entire experience is really special so good um we're working with a neon sign company here in Austin oh really uh to get the like this gorgeous retro neon sign um are you gonna have something similar to uh like in and out's massive you know arrow pointing towards the restaurant yeah except for just pointing people in the direction of the portals.
47:25Yeah, exactly. That same feel, right? We want it to feel like something that is just like so much joy and so much craft went into each piece of this experience. Like someone really cared in every aspect of the customer journey kind of thing. Yeah, totally. I mean, like down to like product photos. I want Peyton to get his day in the sun when he releases this, but like product photos are like, there's something you don't realize wear on you when you use an app. Like if you go and you look at product photos, they're necessarily evil. They need to be like, you know, high contrast. You need to be able to be like, there's so much information packed into the label of a product.
48:07Yeah. Or the way that it looks, that you need to expose all that information to the user. So you really can't get away from them. And they're just like, product photography lighting is like really harsh. and all the different products have a different brand identity so they all end up looking different and just on a phone it's like it takes a lot of kind of like eye strain to like look at it correctly uh he found this way to make them look gorgeous and it's just like such a joy to swipe through and look at all the products um is it like completely different than how you I've never seen anyone getting anything close to this it's like completely different so the tough thing about like doing a new experience like this is it can't feel new which is tough like there are some things we talked about like people have to just intuitively get it yeah yeah yeah they need to intuitively get it which means you need to keep it as close to current experiences as possible but just the most joyful version of those new experiences um or those old experiences so like uh what i wanted to do is i didn't want to charge people until they came and pick up their groceries i just wanted you people put things in the cart just throughout your day especially if you like have a partner right yeah and all of you you guys can just put things in the cart and then just next time you're driving by the portals just go and pick them up you don't have to like push order just like we're gonna have them ready if they're in their cart but we'll charge you wait so it'll just like recognize that you're in the parking lot and then yeah just like prepare your order as is right then yeah and you don't even need to technically click that's what i want to say add to cart and then when shows just whenever you're ready to pick it up just like that will be sick and i do think it's like a a better customer experience yeah it's one of those things where people are confused as hell the first time but then like the next time is this one of those well we're not going to do it because like you know and i i wanted to do it and this is where like uh me and canon um canon just balances me so well like we have a really great partnership um he's perfect uh do you kind of get super excited about things and then he i don't know not necessarily reigns you in but like gives you some good perspective on is it possible to actually execute the thing
50:14sometimes um i want to say that's like that's not quite the relationship but more times than not maybe yeah um but it's like you can't get that's too new yeah people want to feel comfortable hitting checkout that's just like what they're used to so like that is off the table um but the way that we solved the product photo um problem and also there was like you know um getting product photos is all the founders that we've talked to in the space of um grocery delivery is just like getting the product photos it's just such a nightmare getting them good like uh you found such a good way to make those product photos beautiful to get them easily to make it like a very uh homogeneous i think if i'm using that word correctly it's like uh everything has like this same like brand feel even though they're so different so um he crushed it like his obsession on getting this uh product uh just perfect looking just as close to a perfect customer experience um and just all of us is like we will move heaven and earth if it means that it will be 1 % better for customers.
51:23And I think what we're doing is building a playbook for all the people who want to fulfill through autonomous logistics with us is like we're creating a playbook that they can use to go and do the same thing. So goods is just our anchor tenant brand, but all the portals that we're making in Austin and when we go to other cities, we want anyone to run any concept out of any RFC or any store connected to the network to be able to borrow those same tips and deliver the same experience. Crazy. Super cool. If you had to guess, like, let's run two different scenarios. One is things go extremely well.
51:58What does that look like? And then another is things go terribly wrong. What does that look like over the course of the next, like, let's say 12 months and two years? 12 months and two years? Yeah. I think it's, I think if anyone was looking at this, they would think, you know, they misestimated how difficult it is to be a retailer and i don't think it would be that because me and canon are obsessed with retailing oh and both of us independently have been our entire lives like um the amount of time even before going this way just like we got into this because we love retail we love uh people buying just like studying just like sam walton jeff bezos those guys um yeah i mean can't canon's the the walton expert so he could tell you anything right we both love costco we can um uh tell you like in any fact and then amazon i've got i love amazon so much like i love everything they've done and like the decision making a lot of how we built the team and run the team is is based on amazonian culture um just like it it's such a noble industry because you're able to take advantage of new technology faster and better than anybody else, right?
53:17Because you're right there next to the customer. You can just, you know, integrate, especially with an engineering culture, your ability to take new technology and integrate it really quickly in just the right way for the customer is better than anyone else because the people making the technology and then selling it to retailers are missing the customer feedback loop that you need to make stuff really good and joyful to use. And the retailers don't understand to tech enough to know to what ends can they like implement it um like what are the edges like pushing it to its its limit and you see that in just like amazon's ability to uh they have created these fulfillment centers that are like alien tech and they did it quicker than anyone because they both had the need um of having to store a lot of things and deliver really quickly um but they also kind of saw what that technology was going to be able to do.
54:11So they were able to bet more than anyone else was going to bet on it because they knew what it could eventually be. And it's just like that blend of things you don't find unless you at least have a retail arm. I think one of the ones that gave us a lot of confidence was looking at Ocado. Ocado runs a grocery business in the UK. um okato is uh they do warehouse automation um so you've probably seen them it's the robots that sit on the grid yeah on the like ceiling and they just like drop little things and all these different totes yeah so they sell that tech to um grocers but they also run a grocery business in the uk and i don't know if like that tech is so far you know for the grocery business it's like the only tech that's really like exploded yeah um in in terms of automation and i think if they weren't pushing on both fronts like it never would have got there because no one would have ever guessed how valuable it would be um and so so if it goes wrong i don't think it will be that um i think uh you are doing a there's a lot of things that you are doing so i think at any time you're running a business like that like financing it from its infancy to its crawling stage it's just difficult so like that's probably if it doesn't work it's like probably that um it's really difficult time uh growing at the right pace um trying to you know overextending yourself um is there a danger to like scaling too fast i'd imagine that the like ideal situation is you're scaling as fast as possible that you can as fast as you can deliver an amazing customer experience every single time.
55:54Right. Yeah. And I think those, there are people who really understand that calculus. Because like as we're, if, you know, our plan is to put those portals out there, once they're achieving kind of our 42 orders a day break even, then you want it to extend the network, you know, a half mile this way and put portals a half mile this way. um now if it's going really well and it's growing really fast uh those portals should do um just as good the problem is that they're cannibalizing some of your yeah um demographic some people will just start going there um so unlock new areas but you also cannibalize you know yeah especially because like we're building the network they're connected so the distance between the first two stores is different than like if you were um if you're running like a uh chipotle like chipotle in vincidence has one store and you know where were they in the i think they're like san francisco i have no idea i don't know i don't think so anyways let's say they started um yeah in san francisco um they say in the mission like they didn't but if they did there right they wouldn't go and put one like a mile away right they'd be like on the other side of town or somewhere farther away to capture a new demographic.
57:15We don't in the early networks, you're geographically tied to where you're starting. So that second one's going to be, you know, not too far away. Very quickly, though, you start to hit this critical mass where if you are also adding the services on top of that you're offering at those individual portals, your ability to like you know once you have 10 portals um you're you're really close to a lot of people so the value of having something in here and being able to get it out to those portals in under five minutes is huge um so you start to add more things here which means you know adding more uh portals um get you to more people which adds more um you know rfc volume um and those initial portals that are just like a half mile down the road are connected to so much that everyone around there is just like they love going to this portal so they get it the flywheel is great once it starts spinning it really spins yeah and so I think the real challenge that we have is getting that flywheel to that point and what we the way that I see it working is we show people this never before tried customer experience we show how much people love it I always show So, you know, the value of what we think is, you know, I think when we talk about Pipe Dream, you know, the robots are like a really fun piece to think about.
58:38Like, they're really cool. The portals are kind of unique. So, you kind of think about the thing that's like out there in the customer. What we think is going to be our biggest value is actually just the RFC. Like, fulfillment centers are just inherently like difficult to run. And we want to hang our hat on the ability to, in three minutes, get your thing out the door. Just run it extremely efficiently. In whatever the modality. If that's like we're putting in the pipes, we're sending it to a portal, that's how. If we are putting it into a drone, it's three minutes and it's in the drone, it's out.
59:03That drone can get it done in, you know, eight minutes. Why does like three minutes matter so much? Because drones are so fast, right? Like all these other modalities can get things done so quickly. All of retail right now, like even like a micro fulfillment center is fulfilling things in like 10 minutes, right? like 10 minutes is an eternity when the thing that's going to take it to the end point is going to get that done in five right um so we think like uh especially as this industry advances um right now uh there are businesses that are arbitraging on distance to your house right 7-eleven uh is more likely to sell you a coke than walmart is from a frequency basis because are closer to you, it's easier to get, right?
59:52But as autonomy progresses, what ultimately it's going to matter is not, when you're going through the app, you won't be looking at what's closest to you. Even now with DoorDash, you don't look at what's closest to you, look at what's fastest. So making sure that something can originate the quickest and get out of where it's stored matters a ton. And being under three minutes, I think it's going to be an industry standard. and takes like a totally different framework of thinking on how to do that. So that's what we want to hang our hat on. And I think what it will allow us to do is, you know, if you look at what's like a hot consumer, like David Protein Bar.
1:00:32Yeah. I love this. They're sick. When they look at who's, you know, who do we want to fulfill our last mile fulfillment? We want to be able to offer David Protein Bar, you know, take advantage of all the benefits of autonomy. We want people to go pick it up. We want people to have it delivered really quickly. We want them to look at pipe dream fulfillment and say we want it to be there because they're both going to get it out the quickest. They're going to get it close to my customer and get it in their hands as fast as possible for as cheap as possible. So that's the war we're fighting. And so the kind of base layer of that is just you need a fulfillment center where you can get things out really quickly.
1:01:12And then you need to be able to get it as close to the customer as possible really quickly. And then you need to be able to have a method to get into their hands as quickly and as cheaply as possible. And depending on your address or where you are in a city, like, that changes. Like, a drone is just going to crush in the suburbs. You're going to crush anywhere with the backyard. Or in the high-density areas. Yeah, yeah, yeah. It's like perfect. Totally. And as infrastructure changes, like, I think it will crush even in, like, you know, higher density downtowns. I think people over, maybe underestimate how easily you can adapt infrastructure for, like, really breakthrough technologies, which drones clearly are.
1:01:51And so we want to make sure that, like, that is what people are going to be demanding from fulfillment partners as autonomy as well. So we want to make sure that we are, as it is very nascent now, building that capability so that when that happens, we're ready. Yeah. What have been the biggest challenges just over the past year or so? The biggest thing is just like... um how slow even like some of the more startup incumbents are um it's not even necessarily like their fault right it's just like they're massive organizations uh their core business is crushing like anything you're doing that is like frontier just like right now like autonomous logistics is not ai no if it was this would be different right um but even look at ai now ai is clearly massively here and just the ability for incumbents to adopt it is so tough i mean google invented the industry and the they they have struggled so much to implement no not at all maybe just now like it is maybe just now they are like i did gemini is pretty great now um but it is just so tough for them to um pivot especially in um hardware and retail where uh the amount of integration is like an order of magnitude more difficult than adding ai in the loop um and like your iterative process is much slower.
1:03:21Yeah, I think that's probably been, I'm going to say a challenge. It's been a big surprise problem. Yeah, it's just industry players not being willing or able to adopt new technology as fast as maybe they should be able to. Yeah, and like, you know, some of that, you got lucky yourself and be like, that's a self-skill issue. Like, you know, I'm sure we could speed along faster. But then a lot of it, and like, I don't want to talk bad about anybody, but it's just like a lot of it, it's like, God, this is crazy. And then watching like, you know, our industry partners struggle you know people who are you know we have partners who are like fortune 10 and in seeing them work with the same partners and struggle exactly as much as we are um was just like oh man that is um that is crazy like i i think us and and probably other people in the industry see this as such a paradigm shift like autonomous logistics is here it's not even like not here yet or coming like no no one knows when it's gonna come it's here like uh i mean the cars are literally driving themselves cars are driving themselves to get here you know how many millions of drone deliveries are happening every year like this technology is here um and the fact that people are not pivoting their entire businesses on it is bananas the fact that there aren't like hundreds of startups um who want to be like autonomous logistics retailers is crazy well i think i think the biggest thing is like if you have success stories that people can look to then they can kind of say all right i'm going to be the next jeff bezos or whatever but then if amazon just eats the entire you know market then it's a little bit difficult you know you can be gopuff or you can be you know dash mart or whatever but it's really hard whereas you know with ai it's not super hard to just like be on your computer coding up something and just happen to oh i see what you're saying gpt right your point would be like the the amount of risk you take yeah trying to beat the incumbents is a lot lighter when it's software right and so maybe it's like easier to try and then see that people are successful against the incumbents which creates I think there's a lot more there's a lot more case studies or you know examples of people succeeding in software than there is in people succeeding in fulfillment or yeah like you know yeah building a grocery store yeah you know so it's really hard yeah what are you most excited about oh I'm most excited about um creating a great customer experience I think that's uh you know uniquely I think you've asked our partners like I think where we love to create great experiences.
1:05:40Be a good partner is our number one company value. And we just don't get to flex it publicly a lot. But that is the thing that I think more than anything at our core as a team, we care about so much. Do you have a specific story that kind of represents that well? Yeah. When we were boring under one of the roads. I wanted this story to come out some other way. But I'll tell it. There's others like this. But when we were boring under one of the roads, when we started the Georgia project, we were using construction crew that we ultimately fired, and we did it ourselves the rest of the way. And this was, I can't remember if this was when we fired them, or maybe we fired them a little later.
1:06:26But it's not hard to drill under a road. Like you do what's called a GPR. You just like ground penetrating radar. It's like a lot more you walk across the road it shows you where the stuff is you mark it so you know where it is You plan your route and then you bore You just don't not that complicated not that complicated. Yeah our construction crew And we should we should have supervised them more that was 100 % on us Shouldn't have done that But they went and marked And then it rained and the markings went away and they said it But if you think we remember things are delicious for anyways, so they bored And then they hit a gas line.
1:07:08And it's not that bad. It's like they fix it. You know, it's fixed in like four or six hours. The problem is it took out gas for every single business in the strip mall rank. Oh, so just like pissed off a bunch of people. Probably would have. Yeah, yeah, yeah, yeah. So as soon as they called us and let us know what happened, we had someone like around there. And so I called them and I was like, look, right now go to every single business. and you just you walk in you tell them like hey we hit this guest line you're gonna be out we are so sorry um we're gonna give you tell us um here's my number uh tell us how much you make in three days and we're gonna get you that money tonight and so he just went in and like i i think you know 15 minutes after it happened he had talked to everyone and everyone had his number and they all text him how much money they make in a day and we paid him and it was back on that day it was only like down for three hours but we were just like look we impacted you wanted to go above and beyond right and like you know being down for a business sucks like it means that people are coming and they're losing trust that you're going to be open because you just can't serve people and it's like that's a horrible look for businesses so it's like how do we uh fix this problem in a way where people hope the problem happens again that's what we always try to do how do we and when we were you know we had to convince some people here um to let us uh take some of their parking spaces to like build the experience that we wanted and so when we go into it it's like how do we um take parking spaces so well that people want us to take more parking spaces how do we like make so many like give them back more than we're taking so only make the customers happy but also make the people that you're like leasing the property everyone yeah yeah yeah you want them to know that like if something happens you're gonna above and beyond fix it to the point that they're they're hoping something bad happens because they like they benefit so much from it Yeah.
1:08:53Right. We wanted people to be like, I hope they hit another gas line.
1:08:59And yeah. So it's like, I mean, that's just it's just something that like, you know, I think Cannon's one of the most ethical people I've ever met. Like he's just his his his love for people and like moral compass is just like it's so high and it's just innately like. I love Cannon to death like a brother. His girlfriend told me a story the other day. uh sometimes he'll just like go throw her bathrobe in i hope he's okay i'm telling this he'll throw her bathrobe in the dryer and then hang it back on the hook when she's down with her shower so it's nice and warm uh when she gets the shower which is just like how sick is that i mean that's like the right type of co-founder to have oh dude yeah 100 um yeah canon would like lay down on a bed of nails so like it was easier for me to walk on like he's just like really there for you dude he's there for everybody 100 um but it's like true of like everyone on the team is just like if one of us are moving like everyone's showing up to like move them that's just like they are people who care about the people around them um and so with with uh yeah just be able to like take that to like we don't really benefit from that in terms of a flywheel um whereas like amazon really benefits from that in terms of the flywheel uh walmart does like these businesses that have customers where if you treat someone above and beyond it turns into more business like i'm just really excited to like that's the business that you want right to create totally yeah totally and it was one that we knew we were going to have to like even if we work with partners we're gonna have to do that for them in order to like you know make this new experience work i think that's true of any autonomous logistics provider is like you have to think so much for the retailer yeah and eventually we're just like what we're thinking so much for the retailer that we're basically doing the retailer's work we are so close to just like you know being the retailer that like we would have and like people would say no to our ideas so many times or it's like no this is better for the customer we know what it's like and they'd be like no no we don't want to do that and you know their ability to own that entire process early on and prove to people that like caring that much about the customer works and then giving them the playbook to go do that themselves that's a perfect formula yeah has the so i know you said that like when i came last year you were at something like 12 people and now you're still at 12 people like what has it been like to have a super tight-knit team that's all just like super laser focused on making this first store or first you know distribution center happen yeah what's that been like um i'm just like i love you know uh working with a really high talent density team is just the best and it doesn't take many people to get out of that talent density um so that's just something that we care about a lot and we're not always perfect we messed up a million times at it um and sometimes i've been the the you know the talent holding us back and you know sometimes it's been um other people um but it's just like caretaking a very dense group of people who are all really good and it's just like running a sports team right um like you need to run a good sports needs you need two things you need talent and then you need um talent that works well together uh and it just takes one person to ruin that right um like in sports they call it like a someone's like um a uh what's like a locker room toxin right or like they're poison for the locker room it just takes one person uh to get everything off um it just takes one person who misses some threes on a basketball team to make your star player be like, I don't want to play with this person.
1:12:27Yeah. Be less happy. Everyone wants to be somewhere where, you know, like the perfect spot to be, I feel like is the third best person somewhere because you get to learn from, you know, the other people. Yeah, they push you. And then you kind of want to be in that group of people who you're not necessarily doing all the very, very, very hard pressure work, but you're doing that sometimes. You're relied on a lot of times. You're seen as a leader. And you want kind of like everyone to be tied for third around you. You don't want anyone to be worse than you. Working somewhere where someone's worse than you is like, what have we been doing?
1:13:05Right? Like, this is a very valuable time of my life that I'm giving nine to five, five days a week. And then especially if you go over that, it's like, that's an insane amount of your life that you're giving to something. Like, to do that around people who are not as good at you sucks. um so you kind of want like a team of people who are actually the the second best or first best but they all think they're the third best and then they all kind of like look up to each other in that way oh yeah totally yeah a great way to describe it yeah um and we had a really good bar for that early on whereas like we our first eight hires we wanted to we i mean we waited a long time to hire like you know our first second third employee because it's like we wanted each person to come on to be so good that you look at stepping up stepping up to an nth degree like me we wanted to be embarrassing that they joined like for them right like and then you look at like the rest of them be like who's this person working with these people suck compared to this person and uh like we accomplished that um and then we eased off of that um and we shouldn't have and so we had to get back to it why why did that like change happen and then how did you figure that out in course correct because you just end up with so much work that you're like we gotta hire someone this is how it happens every single time it's like you end up you're like okay we're not gonna find this rock star this person doesn't exist you think this every single time and then people especially people outside the hiring process are like we just need someone so these guys look great i'm reading these resumes some of them look really good these guys are really good are they better are they the best would they be the best in the startup would they come on it's like no but it's close and that it's just enough to like lower the bar just slightly and you do that a few times and suddenly you're on a team with like a minuses and b pluses totally like the a minuses is where it kills you um especially when you're looking at a whole group of people and especially when you start to get people who are very well credentialed in there and even though like you do your best to reject it and it's just like there's people that rise above the credentialed people to the point that you mark the credentialed people here or like very high, right, because they're credentialed, even though they're not a fit.
1:15:13You know they're not a fit. But like the danger for us has been like you mark the credentialed people here and then there's someone like this no-name person who's like a little better than them. And what you got to do is look at them against the team. That's the barometer. But then you start to be like, we've obviously found all the people we're going to find. And they're like really, really credentialed. And this person's better. This is the best person I'm going to find. They're like clearly better than the well-traditional people that is good enough and that is where we failed the most To the point that like we're looking for a growth person right now and we had someone who's like Killer and it was just like but they're not better than our bar Quality for quality our bar for quality right now is the best person on the team right now Especially on that end would be koozie She was like top of her class at warden she ran northeast super rides she's had like and she is unbelievable she can just do literally every single thing you need and it's like the people we were looking for growth came from like all the places you want a growth person from but they weren't better than Koozie they're just like we're just gonna run without a growth person till we find that person how long did it take to find her uh months months five five months But as soon as we found her, once you find that person, it's just like very first time meeting them.
1:16:38You're like, that's it? Yeah. Finding those like diamonds, you never second guess a diamond. Ever. We've never, when we found that person, the first time meeting them, we're like, yeah, we've got to find a way to work with them. This is a great person. Yeah. Yeah. If it's ever a question, it's always been kind of wrong. There's been a couple anomalies to that where we brought on someone who is not better than the rest of the team, but then they worked their way into it. But the anomalies have not been worth the ones that didn't. Yeah. Yeah. Do you have any other awesome stories that I just wouldn't think to ask about?
1:17:20So we've always had this roll open called the wild card roll. and we had it open for like a year and a half and just a lot of people apply for it and it's just like okay there's no reason to hire as well what we want is just someone who's like an animal right just just unbelievably good at everything and we you know our thesis was if you can find that person you probably won't have them long but like the amount of time you have them is gonna be worth it and uh like we just lucked into Ahmed so much um someone had recommended that he go work for a company like us and so he applied and I didn't see his email at first but then I was reviewing some emails and I found it and just like read his um his pitch for why you should work and I was like this is the guy oh my gosh this is the guy this guy is insane um dude have to and then like oh my gosh like i was so grateful like ahmed helped with so many things but like when we interviewed him he was like ready to rock and he had come with like in classic ahmed fashion he'd like a hire he had come with all these ideas and he'd hired someone to render them really high quality so he came with like this really well curated deck of like his ideas of like what he like um my favorite thing about him is like uh he came with all these ideas on what we were doing wrong right and how to fix them and that's all the best people have come unprompted telling us what we did wrong like to the point that if someone just walked in you know i don't think this will ruin the the method if someone comes in and walks in and tells us uh this is what you're doing wrong i i think we are inclined to hire them on the spot um if they're this far into the podcast the likelihood that they're a good person is probably pretty high
1:19:03someone's fighting dropping the alpha at the end yeah yeah um uh but yeah just like he just told us like i think you guys are doing these things wrong and here i did this really high quality render of the things i think you can do better and it was just like oh this guy rocks so hard and we were working through like it was kind of like a weird part-time contract we never did a part-time contract before and so we were trying to figure it out and he just moved from new jersey to texas before we'd even had the contract done so he was like there on day one because he wanted to be there they want he is uh he wants to do things right so much and he will go above and beyond to make sure he's doing that thing the right way um and just like his is he is such a clear thinker um on things um but he will like no one's gonna get a cheaper price than ahmed ahmed will his methods for i'm sure he's fine me telling him this but like i've borrowed so much from ahmed like his ability to get contractors down on a price um if we needed something he would go i think i think this is his method he would go on um like google maps and he'd reverse rank service providers by the amount of ratings reviews assuming that the people with the latest reviews were the most desperate for work which is 100 correct and then he would just start at the top and you have his phone out he just hi I need this and I need it for this much can you do it and if they said no he was just saying I'm just going to call the next one just run down the list until someone agreed and then the price he was starting at was here and then he would just keep going lower and keep pushing them all the way down until it was just like unbelievable price and then the person would always work for us for that price it's just like his ability to like push people all the way to what the thing costs and create a great experience anyways um and his like bar for uh the people he works with is so high that if it is not if that service provider did not provide us the best service we are never working with them again and he is like he will never like they are done and then but once you find someone that he you know who just gives that work all the work yeah um guys so you definitely sold me on like going and interviewing him oh dude if someone's gonna be the next elon it's it's him oh well what was funny to me is because i had seen like a few renders of the stuff that he's building and i was just like ah cool it looks like a spec deck you know like great um but then i went and showed up in person and i was like oh it it's here right right here like and then there's like a video of it actually just plowing dirt and i was like oh yeah yeah it's all autonomous totally totally so many people have asked me like you know the like he's you know he says his his timelines are aggressive and people are like can you actually do it i'm just like just you gotta believe in him like if there's one person to bet on uh you should bet on that he's gonna get it done every single time he's gonna get it done yeah i'm super curious i saw him on like twitter maybe a day or two ago where he's basically like i'm creating a list of all the family offices or something like that and i'm just gonna call you know get to get in touch with every single one of them yeah um yeah yeah raise money that way yeah yeah he's an animal yeah um if you if you had to like think about which traits are probably most important to you right now and like that you most highly value agency obviously and uh just like getting shit done and like myself or in yourself and in general like for people that you want to work with yeah agency's there um yeah canon calls it giving a shit i think giving a shit is the number one quality just like caring a lot um because people who care a lot care a lot about all different type of facets right you want an engineer who gives a fucking shit about um about what the everything yeah they want them to give a shit about the customer they you you want them to give a show about performance you want them to give a shit about their workspace being clean you wanted them to like give a shit about um and they're not going to give a shit about everything but they're going to give a shit about things and they're going to give enough shits that um uh they they will catch things right that's the key like pushes you towards perfection even if you can't ever hit it yeah is that part of why you want someone who or you know you're inclined to to like someone who comes in and says i see what you're doing i like it but it's not perfect here's why no i more people who are like i think it's this right like Like, even feeling like you need to prepend something with a, like, I see you.
1:23:38Even though most people do. But, like, you want people who are just, like, so free with their ideas and challenge things. Even if it's, like, socially unacceptable. It's tough to find, you know, those people who also are, like, incredibly empathetic and understand how people, like, view things. But, like, they're there. Right? You can find them. All Mitt's one of those people. um can's one of those people uh it's like a lot a lot of the people at pipe dream are but um yeah exactly um because you want uh you want an engineer who's like working on like the person who's um part of uh running the otter platform uh you want them thinking about like the customer experience of the portal because there's a customer experience of the portal they could create if they change something on the otter that makes the portal do something a little better that creates a better customer experience you want them to pitch it you want them to um see specs for something and just be like, you know, 15 seconds, like, why isn't it 12?
1:24:34Or vice versa, be like, do people care about 15? Like, maybe you should do 30 and we can save this much money because we're making it 30 seconds instead of 15. You want people to, like, care so much that they do the curious exploration themselves and they just can't live with themselves without saying something. That's, like, the key. You want them to fight through, like, any company builds up this political resistance to people challenging ideas. The more people you have who care so much that they're going to fight through it, that creates really great things. I mean, we've caught ourselves a million times because someone's just like, that sucks.
1:25:09This needs to be better. And the nice thing about that, too, is they care about it being really good. And so when they disagree, it's for a right reason. It can be, but they also will disagree and commit, right? It was like, we need to change this thing about Otter because we can create this customer experience. Then it's like, okay, that sucks. And it's all on me to do, but it creates a better customer experience and I care about that. Is there anything else you want to talk about? What do you think we could be doing better outside looking in?
1:25:48Because you're in a unique position where it's like, we've gone deep dive two years in a row. But most of the time you're not here and you're just seeing things online. I think the most important thing is just building the first thing as fast as possible, getting it out there, getting people through the door, and figuring out the things. I think if you have customers and you have some experience, then you can figure out what's working, what's not working, and then you can iterate on it. But if it's all just theoretical, it's all theoretical. And so I think probably the most important thing in my mind, if I was running this, would be number of days until we actually have customers arriving at our portals and they're picking up packages.
1:26:30It may be not even that many. It might be like 10 ,000 items or instead of 42 ,000, it's like, do we actually have to have 42 ,000 items to have a few customers? Yeah. Right. But just working out kinks that way would be, in my mind, the most logical thing to focus on. Yeah, totally. What about you? what's what's the thing 54 days so we open 54 days so this is gonna be like six hours when I left the office this is gonna pack for hours with with oh yeah products oh yeah all this stuff yeah in 54 days to less than two unless we run into like like the city inspection we fail that or like something like that something outside of our control that but X that delay there are people like picking up packages yeah in the next two months yeah we do the whole grand opening and yeah you know people are picking it up so how fast are you gonna to go from you know this location let's say it does does really well you hit your 42 mark per day is it wait wait is it 42 42 a day ironically the uh hitchhiker's guide to the galaxy you know like universe and everything so you need to get 40 that's the answer amazing um how fast are you going to move and to get like the second iteration of the distribution center so we're going through like the permitting process right now is also going to be based in austin yeah yeah just we just start going that way on the network so this is this lies in the network we're building and then we'll just start going that way we'll do one five-mile loop and then we'll do another five-mile loop that way and then we'll do a five mile loop that way so um there's like health metrics it's not so clear as like once we hit 42 clearly we should grow I think there's health metrics that we want to be hitting we want to make sure that we're retaining people like if returning people then sucks right like you know anyone can get people come somewhere something one time but building the habit right we if we are creating the habit and people are using us and our word of mouth quotient like every time we retain a customer how many new customers three months from now is that creating um have you considered offering people i don't know if tours is the right way to think about it but like educating them if they want to learn about how it all works because i feel like that might be a big part of the sales force thing is if people say you know i just toured this like wacky new thing where I can just order and it all shows up in this nice little box and I just drive through and pick it up.
1:28:42Yeah, I was looking at some old GoPuff commercials or like ads that they did and they did that. It was just someone packing an order. So like, I think it was effective because they did a ton of those. Yeah, it's funny because like the best way to sell food isn't by showing like this perfect plate. It's like, here's the cooking process. Yeah. Yeah. And so I think that if you make it really cool and I don't even think you have to have people see the order when they arrive. but if they just know how it works on the back end. So we're going to do that in content. I think content's maybe better than tours.
1:29:12That's a good idea though. Allowing people to come and do a backroom tour. Yeah, we'll try it. The only reason I would say tours, I think content's a great idea as well, but the reason for tours is because the people that are actually coming are extremely high intent. They are the people that care enough, which means you get the early evangelists. That is a great point. All right, we'll do it. All right, that's a great idea. Okay. One more thing. Yeah, okay. um how do you solve the like because i think this is a big piece of it right how do you solve the i there's something at costco that i have to have i can't switch from costco because like i got this thing yeah it's like there's something at like this is actually um we almost didn't do austin because austin has the highest um we think uh austin has the highest um affinity rating for uh one grocer than any other city um hub like austin loves hub to the point that other grocers won't come here or won't like grow here because they're just like they just don't want to compete against that brand is incredible right like incredible incredible grocer um but yeah so how do you how do you solve that problem so the the way that we ended up solving it was in the early days of like discount retailers.
1:30:29There's no way to like gauge like, like, is it actually better, cheaper than my grocery store? Like, I don't know, right? I'm so locked into my grocery store. So they created price matching. Like, we'll just match the price. If we're higher, we'll match it, right? So you at least know that it is similar to the lowest price and then some will be lower, right? That creates the trust that, okay, that actually is a discounter. they actually are discounting. I can trust that. So on selection, where selection matters, we are going to run a program called Multicart. So everything that we store here will be available in five minutes.
1:31:03And then once a day, we'll go batch orders from other places and bring them here and put those products in. So if someone right now is going to Costco and HEB and like a specialty grocer, we'll go get those specialty items for them. And then they're just picking it all up in one place. Wait, so you're kind of pulling like the early Amazon where, you know, it says like we have two million titles, but then like some number of those titles, they have to, you know, pretend order 10, you know, nine books that aren't even in stock. And then one of that very esoteric item, have it all shipped. The nine don't.
1:31:38Did you know about that early Amazon story? Yeah, totally. And when we came up with that, we realized like, oh, that is really similar. But I think it's just because that's what you have to do if you want to create the best customer experience. But I also think that people will just love it. I think they'll find it hilarious. I think they'll love it, dude. You know? The more that they enjoy it, the more they'll talk about it. Like, I think it's, I think it, what's that, like, Sam Altman line where you have to get, you have to make something so good that people will spontaneously tell their friends about it.
1:32:02Yes. And unless you hit that. That's one of the health metrics that we need to get right, right? We need to create something so good that every time someone comes, it creates two new people to come. Yeah. If we can nail that, right? If we can nail one to two, then that's really all you need. If you have, like, some form of retention through there, then you got it. So I think that is like the ability to go get, you know, organic stuff from Whole Foods and then pull in, you know, bulk diapers from Costco and then do the rest. You can just pick it up in one place or have it delivered all together.
1:32:31I think that is like a killer app. And from talking to people, people have loved it. So we'll see. We'll see if it actually is. But that's where we're going to start. And we're just going to keep doing things like that until we're creating the best grocery experience. Because, like, I think where retail is tough is if you don't have the best experience, you die. Yeah. Because everything else is so difficult. So if you can achieve the best experience, it's still difficult, but it's possible. And then if you can fight through that difficulty, then the upper bounds on a retail business is almost infinite.
1:33:02It's one of the best businesses to run if you can make it work. So that's the bet. But yeah, that's how we ended up solving the key item issue, which I think is the biggest issue for switching to different grocery stores. just time constraining it and saying like hey if you want these items you have to show up at this time you know 5 to 6 p.m or something like that yeah yeah and i think what all will are you just gonna have like a part of this this distribution center that's just dedicated to storing those items yeah like just an order yeah yeah um i guess it's over there and i imagine like if you're doing that then you only actually need enough space to store 120 orders right maximum right a little more it depends on what your rate is but you want to hold it for like a weekend plus because you want people to be able to come whenever they want I have to ask Ken, I think the max days we'd hold it would be a week I mean most items yeah there's like some ready to eat things that you just like got a decent day we'll have full cold chain and we'll have freezers and refrigerators so um however long something will last in grocery store will last in here cool um but yeah i think the like the the secondary benefit of that is you're creating you know two types of citizens in in our ecosystem ones that are like freakishly available and then some that are going to take a little longer um and i think what you encourage people to do is like these other retailers You know the kind of they have an option.
1:34:35This is similar Amazon did You know if we start doing a lot of volume out of those stores There's a high incentive for them to just work with you move their inventory in here Yeah, and then we can do that fulfillment for them and offer that same value because like there is the Merchandising capabilities that those retailers have are like that is they're so good There's some things that incumbents do that like you can't touch like we cannot touch someone's merchandising That merchandising they do, finding those products and getting the customers, just like incredible work. So we don't want to like copy that.
1:35:10We want to enhance that and make it as good as possible. So I think that like handover from those retailers to working with us will be pretty smooth.
From the publisher
My second interview with Garrett Scott, co-founder of Pipedream. Pipedream just announced their first Rapid Fulfillment Center, their new autonomous delivery robot Otter, and the introduction of Goods Grocery coming this fall.
