A former Grubhub product leader describes joining Seamless around 2011, managing the merger integration, consolidating duplicate tech stacks onto Java, and building the data science team that created dish-level recommendations. He left in early 2016 to move to Los Angeles.
Chapters
So, you know, going back to Grubhub, that's a pretty incredible marketplace. You know, could you share a little bit more about, you know, maybe what time period that was at? I know you did briefly, but a little bit more about, you know, what you're working on there and maybe some of the challenges of it as a marketplace. Sure, yeah.
So I joined Grubhub in the fall of 2011, I think it was, and— or 2012, I should say, 2012. And at the time, like Seamless and Grubhub were 2 separate companies. So Seamless was very, very strong in New York and had not been successful expanding to other markets. Like we were in other markets, but not, we didn't have big market share or brand awareness or things had just not gone super awesome. So the core of our business was, you know, consumer food ordering. And then we also had a very healthy corporate ordering business. So many of the banks and investment firms, you know, Goldman Sachs, American Express, those types of companies had corporate ordering accounts and they would provide lunch benefits to their employees. So that was actually like, that made a lot of money for the company. I joined to help kind of rebuild the product team. So first few months, just doing a lot of recruiting and hiring and, you know, trying to basically hire 5 product managers. And then pretty much literally right at the point that I finished, that was the point where we like announced that we were merging with Grubhub, who was the other big player in food delivery. There was no Uber Eats at the time, and there was no DoorDash yet. And Grubhub had the opposite problem. So they were very strong in other markets through acquisitions they had made, particularly in like college towns. They were based in Chicago, so they were big there, but they had not gotten a foothold in New York. So they had tried and failed to IPO the previous year, and they failed basically because they couldn't show that they had a plan on how to get into New York, where food ordering behaviors are 5 to 10x more frequent than they are in any other city in the US, because basically no one in New York cooks. It's actually cheaper to order delivery food than it is to buy groceries there. So You know, at post-merger, like my focus changed to figuring out like how we were going to like bring together 2 platforms, 2 businesses that were basically the same, like up and down the line. Like, you know, we each had our customer data, we had our delivery map data, we had our menu data, we had our restaurants, but we were basically in all the same restaurants written in 2 different tech stacks. So Seamless was unfortunately like a .NET setup, which made recruiting really hard in the US. There aren't many like .NET developers in the US anymore. It's not really considered like a modern, interesting language for most, most engineers. Grubhub was slightly better as Java, but we didn't need all this duplication. So I spent probably like a year and a half working on a project to basically like migrate everything over to a new shared tech stack that was based in Java. And then rebuilding all of our products across web and iOS and Android for both brands. to connect to that new set of data. And then we also rebranded and reskinned everything. So it was a long project, a lot of work. We were doing a ton of hiring during this time as well. So it was super duper busy. But we were also able to introduce a lot of interesting new features. We built a data science team and we started doing dish-level recommendations so that if you ordered chicken pad thai and liked it at this one restaurant, like we could identify other chicken pad thai dishes that you would probably like from other restaurants. Because one of the, one of the drivers of frequency is not letting customers in food delivery like just fall into ordering the same 2 or 3 things over and over again. It's very easy to do that. Like, you know, all, all these sites have like easy reordering. And sometimes you just want to have that grilled cheese sandwich. But if you only, if people only order a grilled cheese sandwich, like eventually they will just kind of stop using the platform altogether because they get bored and they don't want to explore. So one of our focuses was on how do we, how do we try to like let people reorder the things that they like, but also try and expose them to new things that we think they would like as a way to sort of like incentivize more frequent ordering. We introduced food photography, we revamped our ratings and review system. So we did a bunch of interesting and cool things, I think, that improved the product. But also we were trying to like get ourselves set for the future, which was basically being able to develop features and functionality for both brands only one time. So we got that done probably in early 2015. And then in March of 20— January of 2016, I left because I decided that I was going to move to LA. I didn't like commuting from New Jersey to New York where we had moved. So I'm the only person that ever moved to LA thinking that he would get a better commuting time.
