Max Mullen: How Instacart Solved the Problem of Grocery Delivery

21:07People were much more willing to participate in e-commerce. Amazon was doing Prime and that was in the zeitgeist. And we had smartphones, mobile phones with GPS and longer battery lives. So shoppers on our supply side of our marketplace could be effective just using their own personal smartphones. It was really possible to make a magical experience. One where you open your phone, pick a few items that you want to order, go through checkout, put your credit card in the first time. And then magically, those things appear in an hour at your house. I mean, that's a great experience. And I think the novelty of getting your first Instacart delivery is something that helped us grow a lot.

21:44You know, that's really where a lot of word of mouth came from.

21:46Mike Maples, Jr.:This is shifting gears a little bit, but I know in the past when we've talked, you've talked about the importance of culture starting from the early days of the company. Did you think about culture from day one or was it something that you started to think about when it's like 20 people or much later than that? How did you think about that? Yeah, culture is really important. I think of it as the religion of your company. And we thought about it, yeah, right around that 20 person mark. We wrote a set of values. Honestly, it wasn't the best set of values. You know, there were eight of them and they were kind of generic, but it was a useful substrate to talk about why people were successful at Instacart and to have in the back of our minds when we were interviewing and to use in rewarding people at the company.

22:29And it began this process of us taking our culture really, really seriously. and eventually we wrote what I think is a set of world-class values in 2016 and then another set in 2021.

22:41Mike Maples, Jr.:How did you pick them? The first rev, how did you pick them? And second rev, how did you decide? Well, the bigger the company, the more people end up involved in a values exercise, but it's a really important exercise because your core values are like the foundation of your culture. And they're the expression of what you believe and the behaviors you expect everyone in the company to uphold over all the other behaviors that are table stakes, like being a good communicator and having integrity. And in the early days of a company, kind of the best way to think about it is you look around at the people at the company that you think are the most successful and you ask yourself, why?

23:19Why are these people so great? Why are they outliers? When they're spiky, what is their spike in. And we looked around and said that ownership was really important to us. That was one of our eight values. And we had one of our earlier engineers as the case study of that. This is a person who would jump on any problem and fully own it until it was solved. And you didn't have to remind them. And they were self-motivated. And so that's an example of sort of how we found one of our values.

23:46Mike Maples, Jr.:Okay. And so why did they change over time? Is it just the reality of just, you got bigger and different and just, you know, there was different ways to fly in formation later than in the early days. Yeah. I mean, I think with values, you update them every two or three years maybe, and you're looking both at what you want the company to become, you know, aspirational values. And then also some of the sort of the ground truth of who you are. And we went from eight values to five values and we expressed them as like unique phrases. and we wanted them to be memorable and simple and useful like mantras, right?

24:24One of the values in the second set of values was called every minute counts. Okay. And you can just picture, you can just say that in a meeting, right? And you can see why. Guys, you know, the company needs to have a high sense of urgency. Every minute counts. Like how can we get this project done faster? What were some of the other ones just out of curiosity? Yeah, we had a value called this is your baby, which is about ownership and taking you know and taking the company um and prioritizing you know your work um making sure that projects get done and treating them you know with a high degree of of priority and then we had a value that's still around today called go far together which is our value about teamwork um as in the proverb if you want to go fast go alone but if you want to go far go together and it was the idea of of teams that succeed and do wildly ambitious things at instacart are ones that have each other's backs that play for the front of the jersey right that are there for the company and the team not for themselves individually primarily and um in this idea of a balance of being kind and passionate right you're like really hard working you're doing your best work um but at the same time you're doing it in respectful way.

25:37And that was a juxtaposition of two ideas that we thought we had kind of nailed the balance of uniquely as a company.

25:43Mike Maples, Jr.:Yeah, I like that. It reminds me a lot of Naval saying, play long-term games with long-term people. But that's kind of a more, the way you're describing it's a bit more grounded, less academic game theory sounding, right? It's like normal, normal speak. Were there any major decisions that you had to make in the early days, apart from pretending that you were in Trader Joe's? Were there any major decisions where it was pivotal in your success? If you decided to go left instead of right, it might have been different. The first one I think about is when we were becoming a marketplace and we were signing deals with retailers and we really had to choose whether or not we ever wanted to be vertically integrated because retailers' biggest fear was that we would end up competing with them.

26:35And this for many years was a reason why some retailers didn't want to work with us. And so we committed to them and recommitted to them that we were only, we were fully aligned and we were only trying to help them grow their businesses and only through them growing their online businesses would our marketplace succeed. And that we would build a fair, a fair and balanced marketplace so that all the retailers that we partnered with, some of whom were competing with one another, would have the same playing field. And that neutrality was important, I think, as part of our decision. So unlike some of the other players, we don't have our own vertically integrated stores.

27:10We don't buy items from manufacturers and sell them directly to consumers. Everything we do goes through our retailer partners, every single order that we shop.

27:18Mike Maples, Jr.:And you never allowed yourself to create any channel conflict around that. That's right. Yeah. Channel conflict is like the most accurate way to say this. We didn't want to have that. And the layman way of saying is we wanted to be able to look retailers in their eyes and say, we are aligned with you. We want to help you. We want to be friendly and we're never going to compete with you directly. So that was important early. Were there any other things that you did that were just pivotal to success where at the time it just wasn't obvious, but you had to make the call? Well, this one was a little bit more obvious at the time, but there was kind of an existential moment around profitability.

27:55That's kind of famous if people worked at Instacart. Right around 2016, we were sort of running the numbers and doing the math. We're growing quite fast. We were launching new markets and that was one of our primary vectors for growth. We would launch in a new city and do a lot of marketing and acquire and onboard a lot of shoppers before there was demand. So as you can imagine, this is a money losing strategy for some short amount of time. And then over the longterm, if that market turns to be unit economic profitable on its own, it's a very good strategy for growth. But as a company, the whole co was going to basically run out of money at a certain point if we kept growing and kept delivering every order or many of the orders unprofitably.

28:37So we had this rallying cry internally, this wartime moment where we said, hey, everyone in the company needs to be focused on this. We really have to make our unit economics turn positive or at least neutral. And it isn't about finding a$5 win. It's about finding five cents or one cent, right? Every single penny mattered. We sort of took one of our values. Every minute matters. And sometimes I said every cent matters. And we went around the company and found each team and what they were responsible for and made sure that they prioritized taking costs out of what they were doing. So just to give a real simple example, we were not charging consumers, not passing through all of the bottle deposits that we were paying at the point of sale to consumers.

29:22So we made sure we fixed that with a better catalog information. We got really accurate with sales tax and other fees and things that we were paying and make sure that we were charging correctly for those things. And then we did some bigger things like we made our algorithms much more efficient so that they would route orders to the closer shopper instead of the further away shopper just to save everyone time and save the company a little bit of money on each delivery.

29:50Mike Maples, Jr.:So marketplaces have lots of interdependent elements. and a lot of times they don't all work together at the same time. So how did that work for Instacart? Were there, I mean, you're kind of gesturing at it, I suppose, with what you had to do with Trader Joe's, but were there bottlenecks in the system that were particularly difficult to remove? And how did you go about that? Yeah, I mean, we have a really complex marketplace in terms of having demand and then kind of two kinds of supply, right? retailers and their catalogs, as well as shoppers and their availability. And so you're kind of spinning a lot of plates at the same time.

30:30You're making sure you have supply to serve demand. You're making sure you have enough demand and it's sort of at the right times of day and days of week and in the right locations, because this is all constrained by geography. And then there's a lot of bottlenecks involved in, in, you know, making one order get delivered correctly, right? The, the shopper has to be dispatched correctly. They have to pick the right items in the store. when there's a replacement, which is often because stores are often out of things, you know, they have to make a correct replacement. And, you know, then they have to deliver and the customer has to be home and all these other things.

31:02And in each of those human tasks, there's an error rate, a small human error rate. And in a marketplace that's this complex, the compounding effect of lots of little potential human errors leads to, you know, some good number of orders being affected by potential human errors. And so you're kind of constantly measuring these things, working through ways to prevent those human errors. And then that translates into a higher quality of service for the consumer. So a good example is we didn't use to scan the barcodes of the products. We just let people say, I found this thing and then I put it in the cart.

31:38And then eventually we realized we did an A-B experiment that shoppers, even the good ones who scanned barcodes were much less likely to make a mistake. Because, you know, think about a can of soup. There's a lot of products that look almost exactly the same as one another, but are slightly different, just the wrong flavor. Or one's the low fat version and one's the regular version of take something like almond milk. And so we started asking shoppers to scan every single barcode. I don't know for sure, but I think as far as the marketplace, we might scan the most barcodes in the United States every single day.

32:07And that really helped reduce that error. And so you go through the whole business, the whole order lifecycle, each step at a time and work a way out to remove the errors from that step. And at the end of the day, you can look at something like the completeness of orders or what we call perfect orders and try to get it as close to 100 % as possible.

32:27Mike Maples, Jr.:Okay. And I suppose, you know, that there's probably a book that you were familiar with at the time, The Checklist Manifesto by Atul Gawande. Yeah. And I remember one chapter in the book, they talked about how pilots at the beginning of a flight, no matter how experienced they are, they just go through a certain checklist and it's sort of just like not optional, right? And you just, you know, you want to do the pre-flight check of all kinds of different things. Even if you're pretty convinced it's all working, you just want to verify for yourself that it's working exactly the way you said it would every time the same way.

33:01Yeah. And checklists are a really important tool. I happen to be a pilot. I don't know if you do that, Mike. No, I didn't know that. And we do use checklists. It's very, very important. And it's not necessarily because you're going to forget something, but because you want to go through the flow that's the same each time of checking all the same things and also the physical nature of that flow, right? I'm touching this button. I'm touching that switch. I'm pointing to that thing. I'm saying out loud what we're going to do in the case of an emergency in this type of situation. It's a muscle memory.

33:27Mike Maples, Jr.:Yeah. Yeah. You know, and the book also talks about surgeons and how important it is to use a checklist to make sure you don't like forget a step. And so, yeah, it's a similar mentality in an operationally complex marketplace. You are applying the same playbook, I would call it, to each step of the order lifecycle. And then things drift, right? Humans make new mistakes or some software engineer introduces a change and that breaks the metric or changes its definition. So you're constantly kind of looking at all of your metrics, asking the question, why is there an anomaly here? And how can we reduce that?

33:59And it's a really important thing that we were a very analytical culture where everyone had both access to the data that they needed to answer any business question and sort of the training and motivation to go independently answer those questions for any project that they were working on. We were like, as a company, super experimental, like ran a lot of experiments and we're very sophisticated about the way that we run those experiments.

34:24Mike Maples, Jr.:You know, we haven't really talked much just in general about what you learned as a founder. You know, what are some of the things that you learned about what it takes to show up as a great founder? And where do you think were your strengths and weaknesses? And what do you think others can learn from your story? I mean, we learned so much. I could fill a whole podcast of just the answer to this question. but um one of the first things that comes to mind and just kind of starting from the beginning is you really have to be passionate about what you're working on and it's going to be difficult right you're going to reach times where you want to give up and if you're really passionate it's harder to do that and if you're never really passionate about the idea in the first place you'll you'll certainly give up so so passion and like you could call it founder market fit is i think pretty important.

35:16What are their motivations? Why are they doing this? Is it just to start a startup or are they hell bent on making this idea a reality and they're never going to give up?

35:25Mike Maples, Jr.:And how were your co-founder relationships and what worked well with that and what was challenging about that? Yeah, I was going to say that's another learning. It's really important to prioritize your relationship with your co-founders if you have co-founders. And in our case, we were sort of the divide and conquer archetype. We each had a spiky skill in some area that the other people didn't have. You know, mine was product, Brandon's was engineering, and our purpose was sort of business and strategy and building and running the leadership team. And we, for many years, just sort of went in separate directions, obviously in a coordinated way, and we were always aligned.

36:01But my advice to founders is to just spend more time, to carve out time with your co-founders to build personal relationships, not just work on the thing that's highest leverage in that moment, but just like go off and do a fishing trip or something.

36:15Mike Maples, Jr.:And did you ever have any sort of disagreements about who was responsible for what, or did you have pretty well-defined swim lanes between people? Based on our backgrounds, it was pretty well-defined, but there was also just so much going on in the company for the first few years that there could have been six co-founders and we'd all have stuff to do. It was, just a little bit chaotic. And that was fun for all of us because our personalities are to do more and work harder and win even better. Yeah. What else do you think founders should know from your journey if their aspiration is to have a company worth tens of billions of dollars someday?

36:56I have a theory that there's three kinds of decisions in a startup. There's science, religion and art okay and a science decision is one that has a right answer and you should take advice from other people on these decisions you should take advice from people who have done it before or who sit on the board of a company that have done it and you should just get the right answer and execute just basically ship it okay and then there's decisions around religion right this is like culture and how you want to run the company like who who you really are and there's many right answers to that question right you can run your company like amazon runs theirs or Netflix runs theirs, or you can invent a totally new culture that's really unique to you and leverages sort of the traditions that you and your founding team have.

37:40But you shouldn't take advice from other people. You should look within for questions of religion. And then there's questions of art, right? These are taste, right? This is like the 10 % of your company where you're really innovating. It's not like the login screen where there's probably a right answer to how that should look, but it's like how your product works and what you want to prioritize in your roadmap. And these questions should not be outsourced. You should not take advice from anybody outside the company on matters of art or taste. And so the advice back to founders is like, be a little bit more thoughtful when you're asking other people for advice and ask for advice about the things where there's a knowable right answer and kind of keep to yourself the really important taste decisions in the company.

38:23Mike Maples, Jr.:So, you know, I've heard you suggest that you look for irrationally optimistic founders, and maybe this transitions a little bit now into your sort of investing work. But what do you mean by irrationally optimistic? Like if there was a quantitative measure of optimism when it comes to the founder's self-confidence and belief in their mission, this person would score 130 out of 100. Okay. It's just off the charts because you're never going to find someone who's more passionate about what you're doing than you are. You set the bar inside your company and outside your company for how mission-driven people can be and how excited they can be about the mission.

39:05And you can't be too ambitious, but you can be delusional.

39:09Mike Maples, Jr.:Yeah. Right. You have to be able to come back down to earth after you explain your mission and then explain every milestone between here and there. Right. So the more ambitious you are, the clearer you have to be about how you're going to go from being the seed stage startup to taking over the world. And it's unfortunate, but sometimes founders have a really big mission, but just haven't really done the deep thinking on how they're going to start small. Now you're sort of putting on the investor hat. So what's the latest? Now you're out of Instacart and you're kind of doing phase two. What's next?

39:47Yeah. So I recently stepped back from Instacart and I'm investing in early stage startups full-time. I started a fund called Workshop. And part of my philosophy is that some of the best companies over the next decade will get built in person in San Francisco. And so I have a founder space also called Workshop here in San Francisco. And I work alongside the founders that I back and try to help them with each of these little inflection points in their business, as well as some of their bigger problems, right? And we sit in front of a whiteboard and literally workshop their problems together and try to solve those things.

40:19And so I try to be a value-added investor and one that kind of looks like the kind of investor I wish we had had in the very early days of Instacart. Okay.

40:28Mike Maples, Jr.:What kind of startups are you interested in? What kind of checks do you like to write? How do you like to get involved? Yeah, I'm typically one of the larger checks in a pre-seed round or seed round. There's no company that's too early for me to get involved in. And it just has to be an idea that I understand and a founder and a company that I think I can help with my experience. And are there any particular areas or is it more just, you know, you've got a pretty broad stack of skills, obviously, you know, from Instacart. There's a lot to making Instacart go, right? So I imagine that that translates to a lot of fields.

41:04Yeah, there's so many great things that founders are working on in San Francisco in 2026. And so I tend to invest, like all of us, in AI companies, in AI native founders. But in particular, I've obviously got a lot of love for marketplaces, and I've got a lot of love for consumer services. And increasingly, I'm interested in healthcare and have invested in several healthcare service companies. And by the way, there's another frontier, you know, that I think we haven't talked about, which is robotics, right? And the intersection of robotics and AI. And I think that's a really, really fascinating space.

41:40And I'm also super interested in sort of how robotics could be applied, you know, in both B2B settings. But I'm particularly interested in robotics for consumers. I'll put out a request for startups on your podcast. Okay, let's do it. If anybody is working on virtual or physical pets, right? like robotic pets for the consumer. I want to hear from you. I'm very interested in this. I just think, I think in the future there will be, there will be living pets and there will also be robotic pets in the household. And that's a, that's an interesting future that I want to learn more about.

42:14Mike Maples, Jr.:Okay, cool. I think we got plenty of good stuff, so I think it'll be good. Okay, cool. Thanks, Mike. Cool.

42:25Mike Maples, Jr.:Thanks for listening to the Pattern Breakers podcast. You can follow me on X at M2JR, and I encourage you to check out our newsletter at patternbreakers.substack.com. I'd love to have you subscribe wherever you get your podcasts so you don't miss an episode. And if you like the show, I'd be grateful if you could leave us a review. Until our paths cross again, I hope you embrace the power of thinking and acting beyond the conventional boundaries. It's the people who dare to be different who truly make a difference.

43:00Thank you.