DoorDash’s Zesty App Brings Social Energy to Food Delivery  ## Document Notes ## Highlights - **AI-Powered Local Discovery** - DoorDash launched Zesty, an AI-driven social discovery app focused on finding restaurants. - The app uses AI to curate recommendations from DoorDash data, Google Maps, TikTok and other web sources. Transcript: Jaeden Schafer DoorDash has just rolled out a new AI social app called Zesty, and it's designed specifically for discovering new restaurants. And to be honest, this is actually a concept that I absolutely love. I 100% guarantee they used AI models, AI code models to help them spin out a really quick sample of this idea that they were thinking of. It incorporates AI in how it gives recommendations. And I think that every company should be experimenting now that code is so much easier to generate with some of these really cool products that integrate AI into kind of breathing new Life into a product and help you capture more market share. If you want to try all of the AI models that I talk about on the show, I'd love for you to try my own startup, which is called AIbox.ai. You get access to the top 40 different AI models from OpenAI to Google to Anthropic to Grok to 11 Labs for audio and a bunch of really cool image models. If you want to try out all of those for only $20 a month, so you don't have to have subscriptions to every platform and forget where you generated things and where all your content is so It's all in one central database and spot. Go check out AIbox.ai. I'll leave a link in the description. All right, let's get back into what DoorDash is doing. So I think there's a couple interesting things here. Beyond just being like a search app, you can imagine this is like, oh, go find restaurants. They're kind of pitching this as a social app. So it's originally just going to be available in San Francisco and New York. So it's kind of geolocated. And they're essentially branching out beyond just doing food delivery. They're getting into more of a social and a discovery space. I think the idea is to try to get rid of the need to read a bunch of different reviews or to look up different menus or even, you know, go on TikTok when you're looking for a new place to eat. So once you open the app and sign in, you obviously do this with your DoorDash account. And then you essentially can ask the chatbot for a personal recommendation based off of, you know, what you're looking for. So there's an example they gave where you say like a low key dinner in Williamsburg that's actually good for introverts. And then it's going to find you specific recommendations. It's going to go scraping through a bunch of different places. You could also ask it for something like, you know, brunch spots good for groups or romantic dinner with a vintage feel. So it's kind of all of these things that are it's very hard to for those things just organically in something like google maps um andy fang who is the co-founder of doordash wrote a post On x about this and said that the way that they're getting all the data is that they're getting information from doordash who obviously has a lot of info on all the restaurants google Maps tiktok and some other places but to me this so fascinating, just where we get our data from. Like TikTok is like a really reputable source for getting interesting information, Google Maps, DoorDash. And then they said that they, quote, curate the best suggestions from the web. So they're also just scraping the web. Essentially what they're doing is it's, the app is learning what your taste is and figuring what you do and don't like. ([Time 0:00:00](https://share.snipd.com/snip/65e5e054-a5ed-4757-a230-99c9cc9d850c)) - **Experiment Rapidly With AI Prototypes** - Companies should experiment with AI to quickly prototype features and revive products. - Use generated code and models to iterate shallow prototypes that test user demand and capture market share. Transcript: Jaeden Schafer So I think there's a couple interesting things here. Beyond just being like a search app, you can imagine this is like, oh, go find restaurants. They're kind of pitching this as a social app. So it's originally just going to be available in San Francisco and New York. So it's kind of geolocated. And they're essentially branching out beyond just doing food delivery. They're getting into more of a social and a discovery space. I think the idea is to try to get rid of the need to read a bunch of different reviews or to look up different menus or even, you know, go on TikTok when you're looking for a new place to eat. So once you open the app and sign in, you obviously do this with your DoorDash account. And then you essentially can ask the chatbot for a personal recommendation based off of, you know, what you're looking for. So there's an example they gave where you say like a low key dinner in Williamsburg that's actually good for introverts. And then it's going to find you specific recommendations. It's going to go scraping through a bunch of different places. ([Time 0:01:16](https://share.snipd.com/snip/a6e61258-4b26-4149-98d6-fb0c5c418eec)) - **TikTok As A Discovery Signal** - Zesty aggregates unconventional data sources like TikTok alongside Google Maps and DoorDash to surface recommendations. - Pulling from social content introduces new discovery signals beyond traditional review and menu scraping. Transcript: Jaeden Schafer Like TikTok is like a really reputable source for getting interesting information, Google Maps, DoorDash. And then they said that they, quote, curate the best suggestions from the web. So they're also just scraping the web. Essentially what they're doing is it's, the app is learning what your taste is and figuring what you do and don't like. And then once you come across a recommendation that you're interested in you can save it you can share it with other people users can view and share photos and comments about restaurants They've visited so they're trying to build like a whole social app where you're discovering content just around food and maybe you'll follow people that you really like all of their Food suggestions and all the restaurants that they eat at. And you want to go try all the same places they do. Now, of course, this like begs the question, can they stand out against something like Instagram? Or maybe you would just go follow like a food page in your city where someone would do this exact same thing. And I think having the discoverability in addition to sharing the content does make it beyond that. I don't think they can make a standalone social app but because there's so much to do with discoverability i do think that is a big value add um one interesting thing that they said about This the doordash said they said at doordash we're always looking for new ways to help people connect with the best of their communities we're piloting this app to make it easier to ([Time 0:02:45](https://share.snipd.com/snip/964be8fe-b10d-491e-b728-bd0e44529e74)) - **Discovery Plus Social Sharing** - Zesty's social features (photos, comments, follows) aim to combine discoverability with content sharing. - That blend could differentiate it from Instagram-style feeds by tying recommendations to search utility. Transcript: Jaeden Schafer And you want to go try all the same places they do. Now, of course, this like begs the question, can they stand out against something like Instagram? Or maybe you would just go follow like a food page in your city where someone would do this exact same thing. And I think having the discoverability in addition to sharing the content does make it beyond that. I don't think they can make a standalone social app but because there's so much to do with discoverability i do think that is a big value add um one interesting thing that they said about This the doordash said they said at doordash we're always looking for new ways to help people connect with the best of their communities we're piloting this app to make it easier to discover Great nearby restaurants coffee shops bars and more through personalized search and social sharing. We're excited to learn from early testers as we keep sharing what local discovery can look like. Bloomberg actually was the first one that wrote about this and broke the news on it. I think a lot of people might not want to download a brand new app just for doing this when you can go on Google. ([Time 0:03:26](https://share.snipd.com/snip/60fe3e01-e585-401d-ac18-68e1db976937)) - **Use Niche Tools That Save Search Time** - Try a specialized discovery tool if it scrapes unique, useful data you can't easily compile yourself. - Trust single tools when they speed the search process and deliver accurate, context-aware results. Transcript: Jaeden Schafer Bloomberg actually was the first one that wrote about this and broke the news on it. I think a lot of people might not want to download a brand new app just for doing this when you can go on Google. But I think as we're getting AI embedded more and more into our lives, if there really is a tool that can go and scrape the internet and scrape these data sources in a new and useful way, I do think people are willing to use it. So an example is, you know, they mentioned that they're getting their data from Google Maps, from TikTok, and from the web and from DoorDash, of course, like, a world where you use all Three of those tools to try to narrow down what you want to eat. And many times, one of my, one of the things I, I dislike the most is the amount of time I spend just looking for the right kind of restaurant. And it's because I'm like, I'm like, okay, like, you know, maybe it's like an anniversary and I'm like, okay, like Italian food or something like that. It seems like kind of a romantic place, but it's like, well, we've already eaten Italian food like last week a couple times. So like, I don't really want Italian food. So if I could just simply say a romantic place to eat, that's open at these times, that's this far away from where I'm at, like, there's a bunch of things that I think would very natural Language make it much easier to go and find. ([Time 0:04:11](https://share.snipd.com/snip/3cb3db06-a1ce-4148-9ec7-e606b97bac23)) - **Frustration With Restaurant Search** - Jaeden describes how prolonged searching for the right restaurant wastes time and decision energy. - He values tools that cut the search phase by letting him specify vibe, openness, and distance in plain language. Transcript: Jaeden Schafer And many times, one of my, one of the things I, I dislike the most is the amount of time I spend just looking for the right kind of restaurant. And it's because I'm like, I'm like, okay, like, you know, maybe it's like an anniversary and I'm like, okay, like Italian food or something like that. It seems like kind of a romantic place, but it's like, well, we've already eaten Italian food like last week a couple times. So like, I don't really want Italian food. So if I could just simply say a romantic place to eat, that's open at these times, that's this far away from where I'm at, like, there's a bunch of things that I think would very natural Language make it much easier to go and find. And if it's scraping through all of the reviews on Google, and on TikTok to get responses, like if it's going and getting this from a unique data source, I love that. ([Time 0:04:47](https://share.snipd.com/snip/cb867d57-fd40-43a4-818f-c9e6f648e821)) - **Strategic Expansion And Monetization** - DoorDash aims to expand beyond delivery into reservations, discovery, and social features to capture more dining occasions. - This diversification also creates new monetization avenues like ads and promoted recommendations. Transcript: Jaeden Schafer They're branching out, they're doing some new areas. Earlier this year, they launched a feature that let customers make reservations for in-person dining. And you could even get like in-store rewards if you do that. So they have a bunch of different kind of products that they're making. I do think this is smart. DoorDash is very well known for what it does, but I think there's a lot more room to grow if they want to expand the company. I think there's some really great ways, especially as you start thinking about discoverability. Why wouldn't DoorDash target people that are looking for where to eat? Because many people are going to want to go eat out and they don't want to DoorDash something to their door. So why wouldn't they find a way to get a piece of the action in regards to that? And this kind of goes to where I believe their plan is down the road. And that is beyond just helping you discover. I think they'll start having ads in there and they'll charge fees and they'll monetize eventually the recommendation of different restaurants or advertise or promote different Restaurants in there and they'll be able to make money from more of an ad network. But not in a way that they're creating a completely new product. I mean, this is very, you know, related to their restaurant kind of business. It's just a new way to monetize that. And we've seen other advertisers or other apps do really well in this regard. I think Instacart has a massive portion of the revenue now that is generated from putting ads inside of their, inside of their app so that when you're, you know, doing your shopping or Whatever, you're going to be recommended certain products and Instacart gets paid a lot of money for that. Uber makes a ton of money from in-app ads. So the ads business, I would say do not underestimate that. ([Time 0:06:09](https://share.snipd.com/snip/24b3e09c-157c-49e8-bfd7-7e05e5169c83))