Feature Name: AI Dish Explorer

Product: DoorDash

Target User: Users ordering unfamiliar cuisines (first-time or infrequent explorers)

Product Context

DoorDash offers access to a wide variety of cuisines, including many dishes that are unfamiliar to users. While variety is a strength, unfamiliarity often creates hesitation and decision friction during ordering.

This case study explores an AI feature designed to help users confidently understand and choose unfamiliar dishes, increasing engagement and order diversity.

Problem Statement

User Problem

Many food delivery users want variety, yet their behavior is dominated by repeat orders within a narrow set of familiar cuisines. Research shows that repeat and exploration are distinct behaviors in food delivery, and that exploration is inherently harder to predict and support at scale (Li et al., 2024). When dish names, descriptions, or dietary attributes are unclear, perceived uncertainty increases, leading users to default to “safe” familiar choices rather than explore new options (Zhang, 2022). This results in longer decision times, reduced exploration, and limited discovery of relevant restaurants and dishes.

Why This Matters

• Decision friction discourages users from exploring beyond familiar options
• Users miss out on cuisines and restaurants they would otherwise enjoy
• DoorDash forgoes incremental orders, deeper engagement, and better catalog utilization

The core issue is not a lack of choice, it is a lack of clarity, confidence, and decision support at the moment of choice.

Proposed AI Feature (Solution)

AI Dish Explorer allows users to interact with a dish directly from the menu to understand what it is before ordering.

When a user taps on an unfamiliar dish (e.g., Biryani), they can ask natural-language questions such as:

• “What does this taste like?”
• “What are the main ingredients?”
• “Is this vegetarian or halal?”