By Lea Mira, RTN staff writer - 12.17.2025
DoorDash is testing a new approach to restaurant discovery that leans heavily on generative AI, social signals, and contextual search. The delivery platform has confirmed it is piloting Zesty, a standalone mobile app designed to help users find restaurants using highly specific, conversational prompts rather than traditional keyword searches or star ratings.
Currently available in public beta on iOS in San Francisco and New York City, Zesty allows users to search for dining options using prompts such as “cozy pasta spots on the Lower East Side with outdoor seating under $100” or “romantic dinner with a vintage feel within a 12-minute walk.” The app then generates tailored recommendations based on a mix of AI-driven interpretation, location data, and aggregated content from multiple platforms.
DoorDash positions Zesty not as a replacement for review platforms like Yelp or Google, but as a complementary discovery layer focused on intent, mood and context rather than static ratings.
Based on screenshots and early user testing, Zesty prompts users to describe what they are looking for in natural language. The app suggests contextual filters such as walking distance, atmosphere, or occasion, and then presents restaurant profiles accompanied by explanations for why each venue matches the prompt.
Rather than simply listing restaurants, Zesty attempts to “reason” through the request. Users can follow up with conversational questions like “What do most people order?” or “Are there good vegetarian options?” and receive AI-generated responses that synthesize menu data, customer behavior, and social content.
According to DoorDash co-founder and CEO Andy Fang, who shared details about the project on social media, Zesty aggregates data from DoorDash, Google Maps, TikTok, and other sources to curate its recommendations. This reflects a broader industry trend toward blending traditional structured data with unstructured social content to influence dining decisions, particularly among younger consumers.
Notably, Zesty does not foreground star ratings or long-form reviews. Instead, it emphasizes vibe, relevance, and explainability—why a restaurant fits a specific moment or desire.
Zesty is the latest addition to a growing slate of DoorDash product experiments aimed at expanding the company’s role beyond delivery logistics and into end-to-end local commerce. In September, DoorDash announced several new initiatives, including in-house delivery robots, a smart scale for order accuracy, expanded rewards functionality, and the ability to make restaurant reservations directly through the DoorDash app. Taken together, these moves signal DoorDash’s ambition to become a more comprehensive consumer engagement and commerce platform, one that influences where guests dine, how they order and how frequently they return.
For restaurants, AI-driven discovery tools like Zesty raise important questions about visibility and control. As discovery becomes increasingly mediated by generative AI models, operators may need to think differently about how their menus, imagery, pricing, and social presence are surfaced and interpreted by algorithms and not just by human reviewers.
At the same time, Zesty could offer upside for restaurants that struggle to stand out in crowded urban markets. Hyper-personalized discovery has the potential to surface niche concepts, neighborhood favorites, and experience-driven venues that might otherwise be buried beneath chains with larger review volumes.
Zesty enters a competitive discovery ecosystem dominated by Yelp, Google Maps, Instagram, and TikTok, each of which already plays a role in shaping dining decisions. What differentiates DoorDash’s approach is its ability to tie discovery directly to transaction-ready data—menus, availability, delivery, pickup, and potentially reservations—within its own ecosystem.
If Zesty expands beyond beta, it could further blur the lines between discovery, ordering, and loyalty, creating new leverage points for DoorDash with both consumers and restaurant partners. It also reflects a broader shift toward AI-mediated consumer journeys, where intent is expressed conversationally and platforms do more of the cognitive work traditionally done by users.
For now, Zesty remains an experiment. DoorDash says it plans to learn from early testers as it continues shaping the product. Whether it evolves into a mainstream consumer app or remains a testbed for future features inside the core DoorDash platform, the pilot underscores a clear trend: restaurant discovery is becoming more personalized, more conversational and increasingly driven by AI.

