Building a Customer-focused Grocery Delivery Startup with Max Mullen, co-founder at Instacart.
Previous to Instacart, Mullen was a product manager at Location Labs, and the Founder & CEO of Volly, which allowed users to collaborate with friends to make group decisions. He also worked at Schematic, an interactive agency that was acquired by WPP. Originally from Los Angeles, Mullen studied entrepreneurship and film at the University of Southern California.
You and your co-founders have built Instacart into a very large company quite quickly. I’m sure you’ve learned a lot during that time, but what lesson from this process has struck you as the most useful?
One of the things that I’m proudest of about the way we started the company was that from day one we truly focused on our customers. We knew from experiences with past companies that it was the most important thing. That being in touch with customers can boost your odds of success, rather than the other way around. That not being in touch with the customers is a guaranteed way of failing.
So from the start, we wanted to be very in touch with our customers. We did our own customer service, the three founders, from the very beginning for a very long time. Like literally in a round robin way taking the calls from customers on our cell phones and not only trying to solve the issue they were contacting us about, but also asking them about the product. What can we do better? We figured out what our first few big future launches should be by asking customers via this process.
One product that came out of that process was the web interface. You were a mobile app only for a while. Mobile first, right?
It turns out that what some customers really wanted was a website where they could do the same thing. They could use a monitor and a keyboard and perform a lot of searches. Which is the way that most customers find items in our product. My co-founder Brandon built that as a little weekend project. It quickly became the most popular platform. It wasn’t intuitive that web would be more important than mobile back then.
You didn’t have the name of the stores on your site when you started—it was just the products, right?
That’s right, when we first started out we didn’t have a choice of stores nor the name of any store. We just had products. We had planned to fulfill those products from wherever we could for the cheapest price for the customer. We didn’t show customers which store the groceries were coming from. First of all, we didn’t have a partnership with any grocers. Second of all, we didn’t think it was that important to customers what grocery store their stuff came from. They were coming from Instacart, we thought. What we learned was that people have a big affinity towards the grocery store they are used to shopping at. In fact, not giving customers a choice of that store pushes many of them out of their comfort zone.
When you were small, how much did you worry about making grocery store chains angry or incurring the wrath of these multi-billion dollar companies that famously guard their customer relationships?
When you’re small you don’t have to worry about what might happen. You’re sitting there trying to just build a machine that doesn’t exist. To build the feature, or build the product, or build the service that customers want and that you know will be successful. One thing to think about is what other players in the ecosystem will want you to do or will think about you. But there’s also a 100 other things that are unsolved. How are we going to promote this? Will customers pay?
In our platform there are Shoppers, who go to the store and shop for groceries for our customers. This is a group of wonderful people who provide great customer service, and who know a lot about shopping for groceries in a fast and efficient way. There’s a 100 things to think about with the shopper side of things alone. You can’t let anyone stop you by focusing too much on any one thing early on.
If we had really sat down and thought about the process involved in building and scaling Instacart in the early days and thought, “What are all the things that could go wrong?” I think we probably wouldn’t have started the company.
Was Instacart always your original idea? Did you guys go into YC with Instacart in your head the way it exists now?
Yes. There is no pivot in our history. We’ve always been focused on getting products delivered in as little as an hour. That always started with grocery. It still remains focused on grocery.
Now you’re so much bigger. How do you find your relationship with big grocery stores has changed? How much easier is it for you to get a phone call picked up or a meeting scheduled?
Maybe I’ll answer that indirectly. We used to not have any partnerships. Now we have over a 100. We are part of grocery innovation. Not only do stores answer our call, but some of them call us. What we offer to grocers is really unique. In terms of what’s out there and also in terms of the needs of the more progressive grocers. They understand that this these sales are incremental and that being able to order online is an important part of their strategy to reach customers. If you've got a big grocery store, you have a lot of inventory that already exists. It's a fixed cost base. We get more orders delivered from that store. And ones that don’t clog up checkout lines. It’s incremental business for the stores.
That extra business is incredibly important to them, and also you can imagine many of our orders are more profitable than the average order because the basket sizes are higher. That’s truly business they weren't going to get if they didn’t offer delivery. The smartest grocers make that calculus and want to offer delivery to as many people as they possibly can. They realize that there are people who are going to go in and buy less groceries at once because that process is less convenient to them, and there are also people who are just not going to walk in at all, that they can now attract via Instacart.
What would you say to startups who have a great idea but in order to execute they need cooperation from bigger companies?
I would say it isn’t always possible to get that cooperation. Especially from the biggest players when you’re small. Sometimes it’s impossible. It’s like if you’re faced with this big upward climb around a mountain. It’s not about climbing straight up. Maybe there’s a way around the mountain that’s faster in the long run. In our scenario, our way around the mountain was we just didn’t even try in the early days to partner with groceries. We just did what we knew we needed to do to serve customers. We put customers first. We said, “How can we get the groceries from where they are to the customers?” For us that meant operating in cooperation but not in partnership with the grocers. In fact, we formed great relationships with managers of stores before we ever had a contract with those stores at a corporate level.
It was after that that we started to form partnerships. I don’t take any credit for that. We have an amazing business development team which gets the credit here. We started with smaller partners. People who were going to move faster and going to be a little more agile, and they were maybe a little more open to trying something new. One of those is Bi-Rite Market here in San Francisco, which is one of the first partners we signed and has been a great one.
Another early partner is Rainbow. It’s a co-op, right?
Exactly. It’s a vegetarian only co-op grocery store. One of a kind. Only one location. Actually I live right near it. It’s an amazing store. A lot of people love it and follow it. When they launched delivery with Instacart it brought with them a whole bunch of customers from all over the city. Which was great for us too.
How do you get all of that product data from stores—some stores have 40,000 SKUs—into your system? Do you get APIs for that?
One of the things that makes Instacart a very competitive option for the smallest grocers is that we are willing to work with them to get their data in any way that works for them. Via email all the way to by API. Even for the medium-sized grocers, some of them don’t have an API, but there’s some features built in to most POS systems that allow them to upload the data to an FTP server every night. Then with bigger grocers, many have an API and we ingest that. It’s all very custom integration because nobody has done this before. We have a bunch of best practices and so we’ll say, “Hey for this kind of products we'd like data every hour. For produce, every hour. For personal care products once a day is fine.”
It feels like a lot of your business has been some level of custom integration. When you started it, you were pitching: “Hey. We’re going to go to the store and grab all this stuff and deliver it to people.” That just looks messy. How did investors react to that idea early on?
I’d say in general most investors did not believe in what we were doing. We got turned down a lot—but that isn’t a newsflash. Like many companies, we got turned down more than we got accepted. We did not have everyone jumping at us to fund us in our seed round. We did find some really smart investors with good future-looking perspectives who were able to identify the fact that there was really huge market opportunity here and that we had a unique way of going after this big market. Investors had to get over the fact that other people had tried this in some form and failed in the past, namely the Webvans of the world. If you could get there, you could see that there really isn’t a reason why this can’t work.
For the first year or so of Instacart a lot of people, investors, friends even, told me: “You guys are crazy. You’ll never make this work. You’ll never make the machine work.” Then we did that. Then they said, “Okay. Well you made it work but you’ll never make it scale. You’re just in San Francisco and that’s not interesting.” Then in a year we scaled it to 15 cities. Then they said, "Well you’ve scaled it. You’ll never make the unit economics work." We spent the next 6 or 12 months and we made the unit economics work. People are running out of excuses on why Instacart is going to fail.
Did any of you guys come from this industry?
Nobody on the founding team came from the grocery industry. I think it was good to have an outsider’s perspective.
You’ve worked on a couple of startups at this point. What would you say to young people who are weighing founding a startup, but who can also turn around and get a job at an established company that pays well?
I would say “do what you love.” If you are really passionate about an idea and you have a unique and new way of solving a problem and you want to start a company around that, and the job offer you have is not something that you love? Then I think that’s a good reason to pursue that entrepreneur journey. If you’re starting a company because you think there’s an opportunity but it's not something you're passionate about, that’s like entrepreneur lesson number 1 that I learned. Which is don’t do something you’re not passionate about because it will become hard. Then the second it gets hard, you’ll realize you’re not in it for the right reasons, and you’ll quit.
On the flip side, if you have an opportunity to do something you’re passionate about at another company and you take that job, you can learn a whole lot about that thing or about just working in general. I think there’s a lot of value there. Even before starting this company, I spent 7 years at an interactive agency. I got very lucky because this agency was growing very fast at the time when I was there. We were working on some of the most exciting and innovative projects possible. When I left I had gained all this experience. It was almost too good to be true. Like a training ground for a product manager. I learned all this stuff just by osmosis over 6 or 7 years.
I don’t think I’d have been able to start my first company without having that experience. I don’t think I’d be successful in this company without having a) started that company and b) had that long 7 years of work experience, to teach me how to work with people.
What was the most harrowing day that you guys had or you were like, “Oh Gosh. This could be the end?”
There’s definitely been tough moments and situations which we had to work our way out of. We try to avoid talking about them. They are also not that upsetting. One interesting day that we talk about sometimes and that was an interesting moment for us, I think, in retrospect it’s a defining moment, was in the fall of 2012. We had built this great system and we had customers that loved us and were telling their friends. We’d spent all this time building this system that would accept orders and assign them to people. Spent all this time distributing it and marketing it and doing grocery stuff. Getting people to order. We never once stopped to think about building in a way to turn off ordering for any reason. We couldn’t turn it off.
We get the most orders on usually Sundays and usually in the evening. This Sunday in the evening in the fall of 2012 it was raining in San Francisco. A lot of people don’t want to go grocery shopping when it's raining, so they order from us. One of us looked at our systems and realized we had way more orders already accepted than we had ever fulfilled before on a Sunday, or on any day, or that we could fulfill during the rest of that day. This was not a company-threatening situation but you have a real chance of disappointing 25% of your most loyal customers. Maybe they’ll never come back or maybe they’ll tell their friends not to use you. What do we do?
Rather than canceling a bunch of orders or trying to reschedule them, we got everybody who possibly could shop, to go shopping. We personally left our houses. I think it was late on a Sunday and some of us drove into the city. We all started delivering orders ourselves. Like in our own cars or in Uber or in a Lyft. We just kept delivering orders until I think the store closed. It was like midnight. I think I showed up with someone’s delivery at like 12:15 am. She had been woken up and came and got the groceries and said, “Thank you.” Even though they we were late, we served everybody. We got all the orders delivered in a way that you’d never would have expected would have been possible, if you just looked at the numbers at around 7:00 pm.
That’s a great story.
It reminds you can do the impossible. It’s usually worth it to do it for our customers and just find a way to honor the commitment that you’ve made to them. There’s nothing worse than a company telling you, “You can order this thing but actually we can’t give it to you, or we can’t give it to you at that price, or we can’t give you on this day.”
So did you eventually find a way to turn off the orders?
Yes, when we realized that we couldn’t turn them off, we immediately wrote code to shut ordering down. Then with the site shut down and ordering turned off in a hacky way, we went about the plan I just described. We went to bed about 1:30 am or so and we came back to the office at 9:00 am that morning, and wrote what we now call the capacity system. It checks every minute of the day in every area that we operate in: How many orders are coming in? How many do we have capacity for? It will shut off slots so we don’t take any order that we know we can’t fulfill, or that we know could potentially be late. That system, the availability and capacity system, is run by our logistics teams, which are now 20+ engineers and data scientists.