Methodology

The difference between an AI that hallucinations and an AI that converts is the quality of the substrate it feeds on. We don’t guess at visibility; we architect it. This library houses the evolving methodologies of Probably Genius—the 109-point audits, the Information Gain frameworks, and the Black Box extraction protocols we use to move brands from digital noise to LLM consensus. These are not just white papers; they are the technical blueprints for the next era of search.

The 109 Point AI Visibility Methodology

The Identity Layer Audit: Quantifying the Invisible Foundation

Abstract: A definitive protocol for measuring brand resonance within the Synthetic Web. This methodology introduces the "Identity Layer Audit"—a 109-point inspection across 8 "Inspection Zones" to identify the gap between human expertise and machine-readable reality. By quantifying "Information Gain" and auditing "Black Box" assets, this paper provides the blueprint for moving a brand from digital noise to LLM consensus.

Ref: PG-METH-109
Last Updated: February 2024
Format: Technical White Paper
Reading Time: 14 Minutes

The AI Integrity Standard™

The 85-Point Threshold: Engineering Citation-Worthy Content.

Volume is easy; citation is hard. This methodology details our proprietary 100-point scoring system used to audit content readiness for AI discovery. We break down the five dimensions—including Information Gain and Entity Density—required to pass the "Integrity Gate" and ensure your brand is cited as a source, not just consumed as training data.

Ref:Ref: PG-METH-004
Last Updated: March 2024
Time to Read: 10 Min
Key Focus: Quality Scoring, Citation-Worthiness, The Integrity Gate™

The Black Box Effect

Unlocking Invisible Expertise: Why AI Can’t Recommend What It Can’t See.

For 20 years, your expertise sold itself through presence and relationships. AI doesn’t take meetings—it reads text. This paper explores why the best firms are often the most invisible to LLMs and provides the blueprint for "Unlocking the Black Box"—extracting your hidden knowledge from media, emails, and your own head to build a definitive relationship with AI.

Ref:Ref: PG-METH-005
Last Updated: February 2024
Time to Read:Time to Read: 12 Min
Key Focus: Knowledge Extraction, Media Mining, The "Ferrari Engine" Problem

The Multi Model Methodology

The Agent Architecture: Engineering AI for Scale Without Slop

Most agencies mono-crop their intelligence, leading to generic "AI-speak." This paper outlines our engineering-first approach: a Multi-Model Swarm that assigns specific cognitive tasks to the models best suited for them, governed by a rigorous "Integrity Gate" that prevents low-signal output.

Ref: PG-METH-002
Last Updated: February 2024
Time to Read: 9 Min
Key Focus: Model Collapse Prevention, Multi-Engine Orchestration, Voice Preservation

Entity Identity

Building the Persistent Node: From Keywords to Connected Data.

Your website is just one data point. To truly own your identity in AI search, you must exist as a persistent entity across the entire web. This methodology covers the mechanics of semantic triangulation—using @id connections to link your brand to high-trust institutional data, ensuring LLMs treat your expertise as a verified fact.

Ref:Ref: PG-METH-003
Last Updated: February 2024
Time to Read: 11 Min
Key Focus: Persistent @id, Knowledge Graphing, Semantic Triangulation

The Growth Map

Where Your Revenue Is Hiding — And How to Capture It

Most AI audits tell you what’s broken; The Growth Map™ tells you where to win. This paper outlines our framework for mapping your market’s competitive landscape, calculating the actual dollar value of AI citations, and identifying the "Wide Open Territory" where first-movers can claim immediate authority.

Ref:Ref: PG-STRAT-001
Last Updated: February 2024
Time to Read:Time to Read: 8 Min
Key Focus: Revenue Modeling, Market Segmentation, Competitor Displacement