How It Works
The Search Party data model is the global living map of a brand’s universe inside AI. It is designed to capture every important player—brands, products, people, publications, concepts, and narratives—and show how they interact across AI models.
The model is global in nature, anonymized by default and shared across customers. That means when Search Party identifies an entity like “Nike” or a relationship like “Nike COMPETES_WITH Adidas,” it exists once, consistently, across the entire platform. This shared foundation makes the data model powerful: every customer benefits from the collective evidence of how brands and concepts are treated inside AI, while still having the ability to layer on their own private logic with tags, prompts, and filters.
For enterprises with heightened privacy requirements, Search Party offers private data model access. In this mode, none of their prompt responses are used to enrich the global model. However, they still benefit from the global ecosystem by leveraging the shared entity definitions, relationships, and metrics. This gives them the advantage of global intelligence without sacrificing data sovereignty.
At its core, the data model works like a graph database:
Entities (brands, products, concepts, people, publications) are the nodes
Relationships (competes with, validates, has affinity with, contradicts) are the edges
Mentions and observations are the evidence that connect entities to each other
Metrics like strength, momentum, sentiment, and visibility score quantify these connections over time
Key Benefits
Provides a consistent, global map of brands and narratives across all AI models
Surfaces not just what AI is saying, but why it matters and what to do next
Delivers shared insights from thousands of brands and entities at once
Offers enterprises a private model option with no training on their own responses, while still tapping into global intelligence
Common Use Cases
Marketing leaders: Understand how their brand compares globally across narratives and competitors
Enterprise teams: Keep prompt data private while still benchmarking against industry trends
Product marketing: Track global feature associations and positioning shifts in real time
Communications: Measure how positive or negative coverage impacts AI responses at scale