What is Deeprank
Canonical Definition
Deeprank is a protocol specification that defines the structure and semantics of machine-readable business identity declarations. The core primitive is the Deeprank Profile (formal spec name: Deeprank Selection Profile, DSP), which provides AI systems with the structured information needed to make accurate selection decisions.
Deeprank does not provide a service, product, or platform. It provides a standard. This documentation is the reference authority for that standard.
Scope
Deeprank specifies: the structure of business identity declarations (Deeprank Profile), the controlled vocabulary for standardized terms, the semantics of fit determination, and guidance for implementation.
Out of Scope
Deeprank does not specify: how AI systems should be built, which AI systems should adopt the standard, how to improve AI recommendation quality beyond structured data, or any marketing or business growth methodology.
Why Deeprank Exists
The Problem
AI systems are increasingly making selection decisions on behalf of users. When a user asks an AI assistant for a recommendation, the AI selects from available options. This selection is based on the AI's understanding of each option, which is often incomplete or incorrect.
Current business web presence is designed for human consumption and search engine optimization. AI systems parse this content differently and often misunderstand what businesses actually do.
The Consequence
AI misrepresentation causes harm in both directions. Businesses are selected for work they cannot do, leading to wasted time and damaged relationships. Businesses are excluded from work they are qualified for, leading to lost opportunities.
The Solution
Deeprank provides a standard for explicit business identity declaration. Instead of relying on AI inference from marketing content, businesses declare their capabilities, fit conditions, and exclusions in structured form. AI systems can parse these declarations and make more accurate selections.
Guiding Principles
Selection Over Visibility
Deeprank is designed for selection accuracy, not visibility maximization. The goal is correct matching between user needs and business capabilities, not increasing the number of recommendations a business receives.
Declaration Over Inference
Explicit declaration is always preferred over AI inference. If a business can state something directly, it should. Inference should be a fallback, not the primary mechanism.
Negative Capability Matters
What a business does not do is as important as what it does. Exclusions prevent bad matches. The ability to say no is a feature, not a limitation.
Binary Fit
For any given query, a business either fits or does not fit. There is no partial fit. This binary model simplifies selection logic and reduces ambiguity.
Specificity Enables Matching
Broad, generic descriptions prevent accurate matching. Specific, precise declarations enable it. The more specific a business is about what it does, the better AI systems can match it to appropriate queries.
What Deeprank Is Not
Not SEO
Deeprank has nothing to do with search engine optimization. There are no keywords to optimize, no backlinks to build, no content strategies to implement. SEO is about visibility in search results. Deeprank is about accuracy in AI selection.
Not a Ranking System
Deeprank does not rank businesses. It does not produce ordered lists. It provides the structured data that enables selection decisions. The selection logic is implemented by AI systems, not by Deeprank.
Not a Marketing Tool
Deeprank does not help businesses attract more customers. It helps businesses be correctly understood. A business that implements Deeprank may receive fewer inquiries if those inquiries were previously mismatched.
Not a Product or Service
Deeprank is a specification. It is documentation. There is no software to buy, no platform to sign up for, no subscription to maintain. The value is in the standard itself, which is publicly available.
Implementation
Deeprank defines a protocol for AI business selection. deeprank.ai provides tooling that implements this specification.
Specification Status
Current Version
Protocol Specification v1.0. This is the initial public release of the Deeprank specification. The core structure (six-layer Deeprank Profile) is considered stable. The controlled vocabulary may be extended based on adoption feedback.
Stability Commitment
Breaking changes to the core specification will be avoided. Any necessary changes will be additive (new optional fields, extended vocabulary) rather than destructive (removed or renamed fields).