AI Acceleration Quotient (AAQ)

A scoring methodology for measuring how deeply AI is embedded in an organization's operations. Developed by Bob Gourley at OODA LLC.

The AAQ produces a composite score from 0 to 100 across three dimensions:

• Machine learning and advanced analytics (default 40%) — production ML systems, predictive models, computer vision, demand forecasting, optimization
• Generative AI adoption (default 35%) — LLM-powered tools, copilots, custom fine-tuning, knowledge management, AI-assisted workflows
• Agentic AI readiness (default 25%) — autonomous decision systems, closed-loop operations, digital twins, edge computing, data infrastructure maturity

Scores place organizations into five tiers, from AI-Nascent (0-20) through AI-Native (81-100).

What this repo contains

SKILL.md                    # Skill definition and entry point
references/methodology.md   # Full scoring rubrics and dimension definitions
references/workflow.md      # Step-by-step research and scoring process
LICENSE                     # CC BY-NC-ND 4.0

Using the AAQ skill

The AAQ is packaged as a set of markdown instruction files — a SKILL.md entry point with supporting references. This is the same format used by a growing number of AI platforms for persistent, reusable agent skills. Any system that can ingest markdown instructions and reference documents alongside a conversation can run the AAQ.

We've tested it on Claude (Anthropic), ChatGPT (OpenAI), Perplexity, OpenClaw, and Hermes Agent. It works on all of them. If your platform of choice lets you upload files as context or project knowledge, the AAQ will run there too.

Claude (Projects)

1. Download this repo as a ZIP (Code > Download ZIP) or clone it
2. Open a Project in Claude and go to Project Knowledge
3. Upload all four files (SKILL.md, LICENSE, references/methodology.md, references/workflow.md)
4. Start a conversation and ask Claude to score companies

ChatGPT (Projects or Custom GPTs)

1. Create a new Project or Custom GPT in ChatGPT
2. Upload the four files as project knowledge or GPT knowledge files
3. Optionally paste the contents of SKILL.md into the project instructions field for stronger adherence
4. Start a conversation with a scoring prompt

Perplexity (Spaces)

1. Create a new Space in Perplexity
2. Upload the four files to the Space
3. Set the Space's custom instructions to: "Follow the methodology in SKILL.md for all queries about AI scoring or company benchmarking"
4. Ask your question in the Space — Perplexity will combine its live web search with the AAQ framework

OpenClaw

1. Clone this repo into your OpenClaw skills directory (or copy the files into a new skill folder)
2. OpenClaw will detect the SKILL.md and register it as an available skill
3. Invoke it through your preferred messaging platform

Hermes Agent

1. Clone this repo into your Hermes skills directory
2. Hermes will index the SKILL.md and reference files automatically
3. The skill is available in any connected conversation

Other platforms

The pattern is the same everywhere: give the AI system access to the four files and let it read the instructions in SKILL.md. The methodology is written to be self-contained — it tells the AI what to ask, how to research, how to score, and what to produce.

Example prompts

These work across all platforms:

• "Score the top 8 defense contractors on AI maturity"
• "Run an AAQ analysis on Walmart, Target, and Costco"
• "Pick the largest companies in the pharmaceutical sector and benchmark them on AI"

The skill walks you through setup choices (research depth, weighting, output format) and then runs the analysis using web search to gather current public evidence.

What the AAQ is for

The AAQ is a starting point for competitive intelligence, not a final verdict. It works from publicly available information only, which means it can't capture everything a company is doing internally. It's useful for:

• Getting a structured read on where companies stand relative to each other
• Identifying which dimension of AI (ML, GenAI, Agentic) differentiates leaders from laggards in a given industry
• Framing due diligence conversations with concrete, scored evidence
• Tracking how an industry's AI maturity shifts over time

It is not a substitute for deep due diligence. See the full disclaimer in SKILL.md.

Methodology overview

Each dimension has a five-tier rubric (0-20, 21-40, 41-60, 61-80, 81-100) with specific evidence requirements at each level. Scores above 40 on any dimension need at least one named source. The methodology rewards demonstrated, deployed capability over investment announcements or aspirational statements.

Weights can be adjusted by industry. Financial services might weight Agentic higher because autonomous trading systems are mature. Media companies might weight GenAI higher. The standard weights work well for manufacturing, CPG, and most industrial sectors.

Full rubrics and scoring guidance are in references/methodology.md. The research and scoring workflow is in references/workflow.md.

License

Copyright (c) 2026 Bob Gourley / OODA LLC. All rights reserved.

"AI Acceleration Quotient" and "AAQ" are trademarks of OODA LLC.

Licensed under CC BY-NC-ND 4.0. You can use this methodology for your own internal analysis and share the unmodified files with credit. You cannot sell AAQ outputs as a service or redistribute modified versions. See LICENSE for full terms.

For commercial licensing or due diligence engagements, contact the authors at ooda.com.