Subscribe
Share
Share
Embed
CRAFTED. is a show about great products and the people who make them.
Incredible founders, makers, and innovators reveal how they build game-changing products and companies — and how you can, too.
Honored twice by the Webby Awards as a top tech podcast.
Discover the future of product development, technology, and AI on CRAFTED.
Hosted by Dan Blumberg, an entrepreneur, product leader, and former public radio host.
Each episode of CRAFTED. brings you closer to the founders, innovators, and technologists who are reshaping industries and building groundbreaking companies. Through thought-provoking conversations, you'll uncover the secrets behind their success and learn how you can apply their strategies to your own ventures.
What product trade-offs did they make? What experiments did they run? And what was the moment when they knew they were on to something BIG?
Guests include leaders of top startups and enterprises, such as:
- Gusto founder and CTO (Eddie Kim) on why the right mindset is key to engineering and product success
- Lattice founder and CEO (Jack Altman) on building a startup that unlocks human performance
- Asana head of AI (Paige Costello) on how AI is changing the nature of work and the way products are built
- Moov founder and CEO (Wade Arnold) on fintech innovation and why he’s rebuilding payments technology from scratch
- BetterUp chief product management and innovation officer (Gabriella Rosen Kellerman) on the “Tomorrowmind” skills you need to flourish today
- Kelsey Hightower, a legendary developer and Kubernetes pioneer, on why great software is emotional — and how you can innovate long after going from zero to one
- Flatiron Health CTO (Cat Miller) on fighting cancer with data
- SYSTM founder and CEO (Matt Lerner) on how to grow your startup
- OpenSpace founder and CEO (Jeevan Kalinithi) on how AI and 3D technology are transforming the construction industry
- DoWhatWorks founder and CEO (Andres Glusman) on running impactful experiments and A/B tests
- Axiom Space head of in-space manufacturing (Jana Stoudemire) on why quantum computer chips may soon be made in outer space
- Plus, leaders from BNY Mellon, Red Bull, AppFolio, Gartner, Katalyst, Betterment, and more
- And, special live episodes from SXSW and NYC Tech Week on AI, biotech, fintech
Listen to be inspired and learn from industry leaders in tech, and beyond.
CRAFTED. brings you inspiring stories of entrepreneurship in technology, insight into tech trends that drive innovation in products, and business strategy insights.
CRAFTED. is brought to you in partnership with Docker, which helps developers build, share, run, and verify applications anywhere without environment confirmation or management. More than 20 million developers worldwide use Docker's development tools, services, and automations to accelerate the delivery of secure applications.
Hosted by Dan Blumberg, founder of Modern Product Minds, where he advises companies on product, discovery, growth, and experimentation. Dan is a technologist and a former public radio host. He’s been a founder and led major product releases and growth initiatives at LinkedIn, The New York Times, Citi Ventures, and more. Before getting into product management, Dan led the team behind Morning Edition on WNYC, the most listened-to show on the nation’s largest NPR station. He also hosted and reported on air for this and other marquee news shows at WNYC, Chicago Public Radio, and for national broadcast on NPR.
Listen to CRAFTED. to find out what it *really* takes to build great products and companies.
_____________
Subscribe to the CRAFTED. newsletter at http://Crafted.fm
Connect with Dan:
- LinkedIn: https://www.linkedin.com/in/dblums/
- X: @dblums
- Instagram: @crafted.podcast
CRAFTED. is produced by Modern Product Minds. Learn more at https://modernproductminds.com/
___________
CRAFTED. is brought to you in partnership with Docker. Learn more at https://www.docker.com/
- LinkedIn: https://www.linkedin.com/company/docker/
- X: @docker
- Instagram: @dockerinc
[00:00:00] Rich Przekop: Mark is the cheerleader.
[00:00:06] Everything's going well. And then, uh, I find I have to be the reality of like, well, yes, the trends are going in the right direction, but we have this problem to solve today. And that's how you get better
[00:00:16] That's Rich Przekop and he's VP of Product Management at Wonder, a food-tech startup founded by legendary entrepreneur Marc Lore. Wonder's longterm vision is to be the everything app for food, with people turning to it for everything form delivery to groceries to meal kits.. and more.
[00:00:35] I can order a steak, I can get burritos for my kids, and my wife could have a poke ball all in the same order.
[00:00:41] Wonder is expanding rapidly as it opens stores that, as they put it, feature 30 restaurants in one location.
[00:00:48] Serving and delivering that many types of food, quickly, is an incredible operational challenge. And on this episode of CRAFTED. Rich tells us how his team has built custom software, including a food operations simulator, to support to it all.
[00:01:04] Hey, if we were to run operations perfectly. This is what should happen.
[00:01:08] Speaker: Welcome to CRAFTED., a show about great products and the people who make them. I'm Dan Blumberg. I'm a product and growth leader, and on CRAFTED. I'm here to bring you stories of founders, makers, and innovators that reveal how they build game changing products and how you can too. CRAFTED. is brought to you in partnership with Docker, which helps developers build, share, run, and verify applications anywhere without environment confirmation or management.
[00:01:32] More than 20 million developers worldwide use Docker's suite of development tools, services, and automations to accelerate the delivery of secure applications. Learn more AT Docker.
And CRAFTED. is produced by Modern Product Minds where my team and I can help you take a new product from zero to one and beyond. We specialize in early stage product discovery, growth, and experimentation. Learn more and sign up for the CRAFTED. Newsletter at modernproductminds.com.
[00:02:00] We're gonna chat a lot about operations and optimization, and I just wanna start with, uh, congratulations.
[00:02:07] Dan Blumberg: You just had your third kid. I have three of my own. And so I'm assuming that your household is perfectly optimized as well.
[00:02:14] Rich Przekop: Oh yes, it's uh, down to the T. Everybody follows all directions all the time. Uh, it's perfect.
[00:02:21] It, it, you know, it teaches you about slack. You have to build slack into everything you do. Uh, so there are some lessons and, and having kids.
[00:02:29] Dan Blumberg: Yeah, so let's get into it. I'd love to start with,
[00:02:31] what's the big vision for Wonder?
[00:02:33] It's a very hot startup, very, very well known founder. Um, and it's raised a lot of money recently, and I'm just interested in what the, what the big picture is.
[00:02:41] Rich Przekop: Yeah, so I think. Mark's Big, big vision is to be the everything app for food. So long-term, you're hungry, you need food, whether it's groceries, going out to eat, ordering in, getting meal kits, like you go to wonder and we will provide. Um, that's the, the, the long-term vision. Right now we're focused on stores, uh, and the idea of a Wonder Store.
[00:03:03] It's sort of like a, a food hall. So we actually have 30 restaurants in one single location. I like to de say, Hey, you know. Think of a Chipotle, um, same footprint, but instead of one restaurant, we now have 30. And so the theory is that in the same space and the same number of people working there, we should be able to do 30 x the revenue.
[00:03:25] So that's, I think where Wonder is not just a food startup, it's a tech startup because we're, we're trying to get scale and multiples that only technology can provide.
[00:03:34] Dan Blumberg: Yeah, and I, I should know, one of the big reasons that Wonder has gotten so much attention beyond the Food is it's founded by Mark Lori, who has an incredible track record as an entrepreneur. He founded diapers.com, which he sold to Amazon. He founded Jet, which was sold to Walmart.
[00:03:46] Rich Przekop: he is amazing visionary. Um, and. An amazing salesperson. I don't, there's probably a few people in the world who could raise the amount of money that Mark, that Mark has raised. And it's, uh, so in some ways it's a, it's a blessing, you know, to be able to work with him. he thinks big.
[00:04:02] Like this is not a, oh, I think I have a good idea and we'll see if it makes money. He's like, no, I wanna change the way people get food. And his vision here is that wonder becomes that platform for people to get food. So we're doing, you know, restaurants now we have a B2B line, um, we've acquired Blue Apron.
[00:04:20] And so you can kinda see there's more we can get into. Uh, with consumers and businesses, but the goal would be, hey, if you're hungry and you're human and you have wonder, we can satisfy that. We can get you groceries in the future. We can get you your meals, we can, you know, get you your meal kits and all of that, um, and do that across the country.
[00:04:38] We're only obviously in, uh, east Coast, right around New York, New Jersey right now, but we're gonna be expanding in the, this area and then eventually, you know, go nationwide.
[00:04:49] Dan Blumberg: Yeah, I, I should also note for, you know, full disclosure, my sister works at Wonder. She's
[00:04:53] helping build a lot of those new restaurants on the East coast. I, I get occasional photos in the, the family group chat of a newly opened wonder in New Jersey or Pennsylvania
[00:05:02] that that team building out those restaurants at the speed, I mean, I thought my job was hard. I don't understand how they can open restaurants at that pace.
[00:05:09] Tell me more about you. Have you always been an analytics and operations person? How did you get to this role at Wonder?
[00:05:16] Rich Przekop: I have not, but I, I studied mechanical engineering so I love,building things. I think there's a, and I love that satisfaction of making something, seeing it work. so that, that's sort of where I came from. I, I did consulting. I had a, I was a PM for a little bit after like two years outta school.
[00:05:34] I lived in China and it's sort of like. Kind of reset, came back to business school at MIT and then my first job after that was at Jet. And it was in a product role, but it was working on our order management system. because at at Jet we were, you know, it was, uh, we had our own products, our own warehouses, but we're also, it was a marketplace, so we had orders coming from all different vendors.
[00:05:55] So this was sort of the central tool that managed. All the orders, both directions. Like what, what are customers placing? What inventory coming in? Making sure everyone's getting their, getting their products. And so that was my first tour. I never like chose to go on that path. It just sort of where I landed.
[00:06:11] Um, but I found it extremely interesting and I think what also got me hooked there was really having the data. I love sql. I, it's, I probably should stop doing it, but I'll, I'll spend hours some day.
[00:06:24] Dan Blumberg: you to stop doing it, right,
[00:06:25] Rich Przekop: Yes,
[00:06:26] Dan Blumberg: won't, need to know it anymore. Right? Is it, will that rescue me? 'cause I never really got good with it.
[00:06:30] Rich Przekop: I know, I, I, I hope I, I think it's gonna be useful too. I've done played around with it. It's crazy. Um, what it can do, we, we have a, a SQL test for our product managers, so we always make sure, again, I think, you know, it's not always. A requirement, but especially in operations, we say, look, it saves us time if you can go and get your own data, do your own analysis than to have to go to an analyst and build a report and see all that.
[00:06:53] Um, especially at a startup, I think it just, just go right to the source. And AI easily did our SQL test, so it was like, okay, it's pretty good.
[00:07:03] Um,
[00:07:04] I'm not building anything anymore, but I think I get that satisfaction of seeing in the, in the data is to say like, Hey, when we're all doing everything, you can see things improving. And that, that to me is like the same. Satisfaction of like, Hey, I built this and it's working. It's like, Hey, we have this team.
[00:07:21] Everybody's working really hard and look, we solved this problem and here's the, here's the proof. And I think that's, uh, that's what's really motivating and why I like the ops side of things.
[00:07:32] Dan Blumberg: I'd love to work backwards a bit to understand the products that you are building, but let's start with the customer experience,
[00:07:39] and then
[00:07:40] All of the internal tools and optimizations that you're building and how that supports the end customer or the restaurant.
[00:07:46] Rich Przekop: So customer experience.
[00:07:47] So if you open up our app, you get a list of up to 30 restaurants. So 30 different types of cuisinesBobby, Flay, lime, salt, barrio, madons, Mediterranean goes on.
[00:07:57] The great thing about. Buying or shopping as a consumer on the app is that you can order from multiple restaurants.
[00:08:03] I can order a steak, I can get burritos for my kids, and my wife could have a poke ball all in the same order.
[00:08:09] the goal is as a customer, you should get your meal within 30 minutes.
[00:08:13] Dan Blumberg: So, uh, the way we are able to put 30 restaurants into one location is through our culinary engineering, figured out a way to. Containerized the way we make the food. So even though we have different cuisines, they're all made essentially the same way, and they all come in standard inventory containers.
[00:08:30] Rich Przekop: So, uh, we're able to maximize our kitchen space and we can move cuisines around in our kitchen. So it's not like we have a dedicated kitchen for one cuisine and one for the other. 'cause of the way everything's sort of.
[00:08:42] Cooked with the same methods that we can move things around and optimize the space. So that's how we sort of fit all of these restaurants into one location.
[00:08:50] as a, as a chef or cook working at. At one of our stores, it should just be like any other restaurant. Uh, but now where it gets into, into my world, the challenge here is to support 30 restaurants in one store. We actually have 700 unique. Items being sent to the store.
[00:09:09] And I like to use Chipotle as the counter example. You go to Chipotle, I think everyone's familiar, they only send 60 items. So this is like your beef, your chicken, your corn salsa, your tomato salsa, all the, the sauces. Uh, and most, there's 60 items being sent. So that's like. Manageable. That's not crazy.
[00:09:28] Most restaurants deal with that. But for us, for this to work, we actually send over 700. And that is not something one person can sort of keep in their head, keep tracking their head. And so this is where the technology in my world really start to come into play is we've built our own proprietary inventory management software for the stores.
[00:09:49] And the goal there is, uh, we need really accurate inventory. We really need to know how many. Things we have on hand, especially things that don't sell that much. You know, Chipotle's churning out beef and chicken all day. We have some things like a branzino which is delicious, but maybe we only sell two or three a day.
[00:10:07] So it's super important that that inventory is accurate. We have no, we have like very small margin of air, um, but at the same time the job of the cooks is not just the manage inventory all day. So the cooks are, and I think one of our managers said, this is like half of your job is cooking, the other half is inventory management.
[00:10:25] I, we'd like to make less of your job inventory management, but it is a critical piece of the business. And so we have software there that, uh, does all the inventory functions, but tries to make it super easy for our cooks. Yeah, it still gives us really accurate realtime information about our inventory.
[00:10:41] So that's sort of the first. Um, part of it. And then that gets into our supply chain side. So we are sending all fresh food. So of those 700 SKUs, maybe 10 of them are frozen. Uh, the rest are all fresh. And this gets into the biggest challenge we have is dealing with shelf life.
[00:10:59] I think you could think of this as typical supply chain software. Um, but it's not that typical because we're dealing with very high frequency and, and low shelf life items. We've actually built our own, uh. Forecasting models in our own, we call like our planning software.
[00:11:14] Dan Blumberg: tell me more about why, you know, there've been food businesses for millennia,
[00:11:21] uh, and I'm sure there are many, many, many, you know, vendors that sell all sorts of technology around food businesses. And I'm interested in why you needed to build custom software.
[00:11:29] Rich Przekop: Yeah. I think the first question is where are you going to add value? I. And we initially actually had a, Microsoft Dynamics, which is A ERP, enter enterprise resource planning software.
[00:11:43] So like Oracle, SAP, there's a bunch of these that a lot of manufacturing companies use. 'cause we thought just manufacturing, we can use that software.
[00:11:52] We couldn't get it to work for what we were doing.
[00:11:54] And I think the problem was the frequency at which we were. Making things was just not built into the way those software worked. They're for manufacturing. They're, you know, if you think of a car, they're gonna run a line for months so they can set up and plan and do all of this stuff. Every day we're, we're making something new and we could be making something.
[00:12:15] Every single day we're switching what we're making because we have to produce 700 items. And if you think about, it's not even just 700 items, it's probably at the end of the day, over a thousand different recipes, You know, you have to make a ni marara sauce first before you add it to the pizza.
[00:12:29] And so we couldn't. In the right amount of time, get that software to work for us
[00:12:35] And I think the other lens of this is usually third party software, I'd say it's like, get you 80% of the way there. You know, it's kind of the. 80 20 rule, 20% of the effort, you get 80% of the way there. But we realized, and, and where I'm really focused is on waste and stockouts.
[00:12:52] We have really aggressive goals and the third party software wouldn't get there. And if you wanna get it there, guess what? You're gonna hire consultants, developers to customize the software and by the end of it, you're gonna spend probably more money than would've just to hire a development, you know, a development team from the beginning and.
[00:13:10] we also already had that team in place because we were definitely doing something completely new in the food truck world. So it was at that point it's like, Hey, these are smart guys. They built some good stuff. Let's just apply it to the stores and this way will set us up for the future.
[00:13:27] you mentioned two of the key KPIs for you. One is waste and one is stockouts. The, the, you know, there's no Bruno. We ran outta Bruno
[00:13:33] Dan Blumberg: today, and it, I I, it sounds to me in a weird way that that's sort of like unemployment and inflation. Like you, you can't have zero of either, and they kind of go in inverse directions.
[00:13:42] If you have a lot of waste, then you're not gonna a lot of stock outs, but, you know, um, so there's some optimization there. I know you built a simulator. Uh, and I'd love to hear
[00:13:50] more about, uh, how you built that, why you built that, what it does.
[00:13:54] Rich Przekop: Yeah, it's been a great tool for us. And so the, the history is we had a forecast we could say, predicting how much we're gonna sell of each of those 700 SKUs. And there's like a basic. Sort of inventory model that you think about of like, you have your beginning on hand, you receive inventory, you use it, maybe you expire it, and then you're ending on hand and you can kind of play that out.
[00:14:15] Um, you could do it in Excel. In fact, we were doing an Excel to start and you could play this out and say, okay, well you could identify when you're gonna run out of inventory and say, I need to produce enough so that we don't run out. the problem with that we found was that while it worked, it got us numbers.
[00:14:31] We were overproducing most of the time. And so our, uh, one of our data scientists are actually in operations research guy had this idea that like, hey, we're getting a forecast that has a distribution because, you know, we're not able to predict the future perfectly. So every, every forecast of, Hey, we're gonna sell two brands, Zenoss, we actually have a distribution curve with that, given the, the spread, the, the variability in those sales.
[00:14:57] And so his idea was, well, why don't we just simulate. The business and do like a Monte Carlo simulation based on those distributions. And so he built this simulation that says, okay, like we produce something, it takes a day to get to our, our stores. So we have to count for that. We have sales that play out.
[00:15:17] It'll result in expiry, it'll result maybe in stockouts. And using all that, we can calculate and optimize, well, how much do we need to have? On hand in our distribution center and how much do we need to produce to keep that, you know, to maintain those levels soit not only like helped with our production, And what that means is some of our items will have waste. We want it to be zero, but there's some items that are so volatile, like the brand Zino doesn't sell every day. If it's sold one a day, I could zero waste, zero stock out to plan it perfectly.
[00:15:46] But it goes up and down. And the, uh, idea of the simulation is it gives you that, that benchmark, and it's not zero because that's impossible. So it gives you this realistic. You know, long-term goal. And so that's how we've been able to do a lot of work to improve our waste in stockout. So it's both served as like a functional tool to do our production as well as a metrics tool to give us, uh, to give us goals and to really actually understand when things break, why they broke.
[00:16:14] Dan Blumberg: Yeah, I'm sure there are lots and lots of areas where you could optimize something or build custom software here, and I'm interested in how you
[00:16:22] balance, trying to make it perfect versus trying to get it out the door and ship that first, you know, MVP, I'm, I'm, I'm interested in how you approach that.
[00:16:29] Rich Przekop: Yeah. Well, I think I learned, I learned this lesson at Jet where, um, we had our product and development teams were, um, very siloed the idea was like, your team can just do your thing and focus. And what happened was you over-optimize because I was in the role of order management and.
[00:16:49] Once we're managing orders and they're going through the system and there's not a lot of errors, you don't really need to improve the software. But everyone's natural intent is like, well, I'm wanna make it better. I have more things I could do. And if you think of like an s-curve, right? It's like you're at the top of the S curve.
[00:17:05] You're like, for our all intent and purposes, you should be done to very much diminishing returns. So I've, I've, I've been there and I think, what, what? There's like, that was a big lesson. The other lesson from our food truck days was having extremely accurate data. So one anecdote from those days is, uh, we were measuring order e times, and sometimes we would get really long order E times because a truck, you're, you're sitting in, you know, point A and the truck's like miles away and there was a closer truck to you, but we didn't send your order to that closer truck.
[00:17:40] And everyone's like, why? The algorithm's broken, everything's a disaster. And it came down to, it was like inventory. We, you know, we're pulling inventory. Well, this truck didn't have this one thing, and we couldn't explain that clearly. Like everybody knew it, but we didn't have the data. We couldn't put it in a graph or a report to show, no, guys, this is actually what, what happened?
[00:18:02] And so those two lessons, it's like, okay, we need to measure as much as possible, everything we can. And so when we talk about waste and stockouts as our KPIs now. We use that to prioritize. 'cause we can now say it's not just like, oh, I'm guessing that this thing will improve waste. I could actually look what's causing waste.
[00:18:22] It's expiration and back to the simulation. Well, did we expect there to be waste? If we did, then there's nothing to do. But, oh, well look, there's this other waste we did not expect to happen. That's a problem. And you can kind of keep double clicking into that. And we've built up our products to have all that data, so now we can explain what's going on.
[00:18:39]
[00:18:39] Dan Blumberg: Yeah, I, I, I know you, you built a way to automate root cause analysis of data issues, and in my head I have sort of a product manager asking why five
[00:18:47] times is what the system is, is doing. I dunno if it's, I'm sure it's nothing like that Exactly. But, um, but tell me more about how you, how you built that, how you,tested it, how you, you know, iterated along the way.
[00:18:59] Rich Przekop: Yeah. So that, again, lesson learned from the, the trucks. But we, we started with stockouts.
[00:19:04] And, and just to be clear, stock out is basically when, inventory goes to zero. And so it's no longer available for the customer.
[00:19:10] so we had to figure out is okay if something stocks out, so let's pretend it went to zero.
[00:19:15] Well, why did it go to zero? What are all the reasons? And it's sort of like a bifurcation. There's always two, really two causes. One is sales, so we sold it. And then the other is we messed up in operations. Maybe we dropped something or maybe somebody lost it. Or quite frankly, there could be stealing someone.
[00:19:35] You know, we have really good choice, cuts of meat. We gotta make sure those are staying at the store.
[00:19:39] So we're measuring people are removing it out of inventory or if the sales are removing it. So we could say, Hey, in a day, if we went to zero, did we sell more than our like one or two standard deviations above our forecast? And if so, then we're like, okay, well that was a forecasting problem.
[00:19:54] If we did that and somebody dropped something, well then it's a both. But if we didn't sell a lot and we still stuck out, then now it's an operational issue. And so that's sort of step one. And then I. There's also an issue of maybe we didn't even get the inventory to the store in the first place. And so you could apply the same logic.
[00:20:12] You go back to our distribution center, okay, well, did they have it on hand or not? And then if they did, that's an operations issue at the dish. If they didn't, now we're back to production. And then you go back, did we tell them to produce enough? So is it a a planning problem or did we tell 'em to produce 10 and they only produced five?
[00:20:31] And so you can like follow the chain, but that's a lot of different systems, a lot of different data that you have to. Piece together. And a lot of that work happened in, um, in our data, like in, we use Snowflake and DBT. So like a lot of that was done just using sql, uh, mostly and getting all that connected and then running some jobs so that when it comes to like reporting, it's readily available.
[00:20:53] We've done all that logic ahead of time. And so by the time I look at our reports in the morning, like that's all there and I could quickly see what, what happened and why.
[00:21:02] Dan Blumberg: And, and so you, you've built this all for, for, for Wonder and its operations, but correct me if I'm wrong, uh, uh, potential Next step for wonder would be to. Take the software that you've built and apply it to other businesses, other retail operations, similar to how AWS started out as an internal tool at at Amazon.
[00:21:19] Is that, is that the double secret probation plan?
[00:21:23] Rich Przekop: Yes, it's the double secret plan. Now you're in on the plan and only you, um, yeah, that, I think that's definitely in the cards, uh, personally. Someone who develops software and products, especially in operations, I think that'd be, uh, that'd be a dream. I will sprinkle a little bit of reality dust on that, which is, you know, it took Amazon a long time to get there.
[00:21:45] Um, they had to work through their operations, but I think that's, that's our path as well. Uh, today, It's all wonder. So we're buying our own food, we're making everything, we're distributing, we're doing all of that. We're cooking all of it. So, uh, the interesting thing will be, you know, how do we start separating that out so that we don't have to be everything?
[00:22:04] I think, with the moment you could start plugging in alternatives at every stage. You could say somebody else gives us the food or somebody else's distribution or somebody else cooks. Um, that's when that platform can really take off. And I think. We are developing software under the philosophy of design for that future, but build for today.
[00:22:23] So that's always in our mind. If you ask tomorrow, can we flip this on? And there's APIs and you can connect, like absolutely not. But it's not a stretch, you know? I mean, a year or two if this starts happening, that we're in a good place to be able to do that.
[00:22:36] Dan Blumberg: With that future state in mind, I'm interested in how you harden the APIs, you know, for future external use, or that you worry about that later, or just how you approach, you know, what to ship today with, with also an eye towards, how to iterate and expand tomorrow.
[00:22:51] Rich Przekop: There's two things. There's internal and external factors. I think the internal are, are discipline and,
[00:22:56] we just hired a new CTO and he's comes from Amazon and uh, I think Rakuten. So he has a lot of the tech supply chain experience, and he's like, no, everything we gotta, everything has to be in systems and handoffs and all of that.
[00:23:11] the other way, uh, that I think about it is, is measurement. Um, we wanna start. Measuring things, and if everything's connected, it's sometimes hard to pull out, well, like back to this RCA, well, whose fault was it? And that became obvious is that like, Hey, we don't actually have a good handoff between one team and another, or this part of the process and the next part.
[00:23:33] And if we want that to be clear because we wanna start measuring things, that's another, I think. Forcing function. Um, but at the end, if you, you know, have the eye to like, Hey, at some point we're gonna have alternative vendors and alternative this, or we might outsource this. You can start thinking about that.
[00:23:49] I think the timing's just a tricky question. Um, there's no, there's no right or wrong answer. Maybe one way to answer it is, Hey, you've hit your, your metrics for the quarter or for the year, uh, and you have some, you know, some bandwidth, and that's a good time to start working on it.
[00:24:03] Dan Blumberg: I'm sure the ambition is there. You
[00:24:04] talked about Mark, Lori and, and his ambition
[00:24:06] tell me more about the, the culture at Wonder as you've gone through this rapid expansion and, you know, several iterations along the way, it's, it's never easy for any growth stage startup. What, what's been sort of the hardest part in terms of scaling up from a team perspective?
[00:24:21] Rich Przekop: Yeah, it's, you know, the culture's great to mark. he loves what he does, right? It's, it's very obvious. It's, and it's infectious. And so he wants that for everybody.
[00:24:31] And, uh, cultures and values are, are really important to him. And we actually, it's part of our review process. Um, we bring that into the review process. So I think there's like that sort of. You know, top down. But then in, in our world, and I think what we're looking for, especially where product and operations align are like very analytical people who like solving problems.
[00:24:52] And I would say that's the main, you know, the main thread with everyone working, especially in operations and, and product in the operations world is these are all. Cool problems to solve. Um, and if we're giving you the data, then it's even more fun. Um, and so that's how we kind of position it, is like, look, the, you know, mark is the cheerleader.
[00:25:13] Everything's going well. And then, uh, I find I have to be the reality of like, well, yes, the trends are going in the right direction, but we have this problem to solve today. And that's how you get better. You have to admit that there are problems to solve and, and it's not, you know, people's fault. So I think, um. Part of that job, that culture is making sure we hire people who fit into that.
[00:25:34] Um, but one of the challenges, uh, actually has been, um, when we shifted from the mobile restaurant business to the stores, we had to do a lot of layoffs across the company. And that was really tough. Um, doing them, being a part of that and seeing friends and colleagues, um, that was probably definitely a low point.
[00:25:54] And I think what helped bring people around is that, you know. Reminding them it's, it's for the benefit of the business and that there are still all these things to do. And that, um, what also helped is giving people ownership and responsibility, I think has been a, another big thing. So you don't feel like you're.
[00:26:13] You're just here and at the whim, the company can let you go, but you say, Hey, like you own this metric, or You're, you're responsible for this and expect you to deliver on that and work towards that every day. I think that helps get people, you know, back to the, the culture thing, I think and problem solving that gives them the motivation to show up and to work hard and to feel that sense of accomplishment.
[00:26:37] And, During that transition, it was hard because we didn't have that clear, those roles clearly defined. We didn't have the metrics clearly defined. But now that we do, I think, um, you know, it's been, it's been great
[00:26:48] Dan Blumberg: you're building a very bespoke set of software for a very interesting and unique food operations business. I'm interested what lessons you think, really apply to other builders out there?
[00:26:59] Rich Przekop: The biggest lesson I learned is the power of setting clear, measurable goals. I. So go back to this transition period, all that, what really helped us to say, look, you know, we have all these supply chain products, right? We have, I have products in our recipe management.
[00:27:14] So look kind of, we call it cookbook, but it's what the chefs use to make these recipes. We have products in our production facility to help with management. We have the forecast planning inventory, all that. And each of those you could think of as an own product that has its own KPI. But what was powerful is realizing that they all roll up.
[00:27:32] To two metrics, which are the waste and stock up. And you talked about the, those like Libra scales, like right, if one goes up, one goes down. And being able to bring them both down is, is the challenge. But the way I always tell my team, like, look, if the forecast is completely inaccurate, but those two numbers, waste and stock order are zero.
[00:27:49] I wouldn't care. Now that's not reality. But um, but having that clarity and realizing that everybody, what they're working on, figuring out or understanding how their contribution rolls up to those metrics was huge. And then we were able to, once we have that goal set, targets measure it. And then the second piece was, we talked about earlier is this automated RCA is actually to show like, I can now break down this metric and apply assign owners to it.
[00:28:16] So I could say,PM over here. You own this one, you own this one, you own this one. And it just gives them, I think that clarity and ownership. So they have something that they can look at every day and, and pull up a dashboard and say, how's this metric working?
[00:28:29] I think what I found is that I. Um, the people who are every morning, if they're opening up and looking at a dashboard and say, Hey, I saw this, or going on, like, that's, that's a good sign. I think that's what gets people motivated and I think it's, it's not always easy, especially in operations to, you know, assign those, those goals.
[00:28:47] Um, especially when the software is not the product itself. Everything you do on those apps is measurable, and so it's really easy to set those targets and metrics and measure them. I think the challenge in, in products that are used by operations or in in physical world are making sure you have, you can measure things and get the data you need to be able to set these goals.
[00:29:07] Dan Blumberg: I love it. I, I love what you're building and, uh, I'm excited to see where you take it next. Rich, thank you so much.
[00:29:13] Rich Przekop: Thank you, Dan. It's great to chat with you.
[00:29:15] Dan Blumberg: That's Rich Przkeop,
[00:29:16] Speaker: I'm Dan Blumberg and this is CRAFTED.
CRAFTED is brought to you in partnership with Docker, which helps developers build, share, run, and verify applications anywhere without environment confirmation or management. More than 20 million developers worldwide use Docker suite of development, tools, services, and automations to accelerate the delivery of secure applications. Learn more at Docker.com.
Special thanks to Artium where I launched. Artium is a next generation software development consultancy that combines elite human craftsmanship and artificial intelligence. See how Artium can help you build your future at artium.ai
and CRAFTED. is produced by Modern Product Minds, where my team and I can help you take a new product from zero to one and beyond. We specialize in early stage product discovery, growth, and experimentation. Learn more and sign up for the CRAFTED. Newsletter at modernproductminds.com.
[00:30:12] And please share CRAFTED. with a friend so we can forecast demand.
[00:30:17] Rich Przekop: And everyone's like, why? The algorithm's broken!