A former TripAdvisor product leader explains how restaurant discovery works differently across regions. A "hot and trending" filter flopped in Italy, where diners wanted "family-run" restaurants instead. Success in hyperlocal marketplaces requires understanding what each place actually values.
So kind of going back to your time at TripAdvisor, which of course is a marketplace that many of us here know about, you know, what were some of maybe the kind of key learnings from your time there that you've then taken with you to Sesame?
Oh gosh. I mean, I think, I think there's so many learnings that I've taken to Sesame and like, you know, get too philosophical, but life in general, like it was such, such a great company to work at. It still is a great company. Some of my professional colleagues have worked there, still do work there. I mean, I think the— it is really like an interesting— marketplaces are so fascinating, right? So it was down to like sort of a neighborhood by neighborhood. So especially when I worked on restaurants, neighbor restaurants are hyperlocal, right? A pizza restaurant 30 miles away from you is really not that useful. Right. And so how do you deal with near me? And what is the trade-off between distance and quality? Distance, quality, and availability? Distance, quality, availability, and cuisine type? In the morning, you're not really willing to go that far for your cup of coffee. If you're traveling and you're in Barcelona, it's a once-in-a-lifetime opportunity and you want to go to a a place, well, yeah, you'll travel for dinner, right? So it's contextual. It's contextual based on the scenario, based on the person, based on the event. What we found out about restaurants in general was that there's no personality type in restaurants. There's no archetype really, right? It's really based on, are you looking for a romantic evening? Are you looking for Do you have dinner with your kids? Are you looking for— are you traveling? Are you not traveling? Is this a special event? Are you with a crowd? So I'm very happy eating a hot dog. I'm also happy going to a very fancy restaurant. I'm also happy eating street food, right? But it all depends on the scenario. So really understanding all of that goes into a marketplace, right? So what A good story, quick story about TripAdvisor that I always loved, and I just came back from vacation in Italy, so I was just thinking about this one. So in New York, we launched a filter on our marketplace for hot and trending. And so that would be like the hottest restaurants in New York City, like what are the new restaurants that you want to go to, right? It was really hard to do this because you wanted a new restaurant, but you wanted it to be good. And also like that means By definition, you had fewer reviews about it, which also meant it was more subject to people could game the system a little bit if they knew how to do that. So we always were very afraid of that. But Americans love new restaurants when they open up, like, let's go to that restaurant. We launched that same feature in Italy, and it bombed, like absolutely bombed. Nobody liked it. People were like, why the heck would you want that? And so we did some user research like you're supposed to do, and we should have done it beforehand, which is bad PM, bad PM, like suckers. And what we found is that Italians, they think just the opposite of Americans. They're like, why would you want to go to a new restaurant? They just opened up. They don't know what they're doing yet. They probably just hired their staff. It was like, if a restaurant's in the family for 3 or 4 generations, then they start to know what they're doing. So we launched the opposite filter, which was like family-run, generationally family-run. And that filter in Italy specifically did incredibly well, but did not do well in New York. So it's all about context with restaurants and with marketplaces. And I learned a lot of those restaurant— a lot of those lessons from TripAdvisor.
Yeah, that's a really great story about, you know, hyperlocal marketplaces and how nuanced each kind of market can be on a market level.
Yeah. I mean, like you go to different cuisines, like street food is big in some areas. I mean, I was obsessed with keywords and tags, right? So, you know, you go to Miami, you want a Cuban sandwich. Like, I think if you walk around the halls of TripAdvisor to this day, I'm known as the Cuban sandwich guy because I was like obsessed with like, how can we find that local food that is like down to the neighborhood, like relevant. And that matters when you're traveling, right? So you need to invent some natural language processing and other things to like really understand that. You know, back to Italy, like there's different types of restaurants. There's tavernas and there's oestras and there's ristoranti. And, you know, ristoranti, you actually, you know, that's a really formal meal. You really probably want an oestra more often than not. So you had to really get really nuanced to hyperlocal marketplaces to understand that.
