Theory
I've published.
Seven frameworks for the agentic era — one technical whitepaper on an IAB Tech Lab standard, one cognitive-science measurement model, one psychographic prediction architecture, one operating methodology for international market entry, a seven-layer data collaboration stack, an agentic operating system, and a commercial productization system. All open-access.
Where to start.
Pick the one closest to the decision you're making this quarter. Each framework is linked to the advisory work it's most often used in. For the single named tests and terms coined in the essays, see the glossary.
- Start with EARO if you care about measurement in the age of agentic AI.
Used in the Attention playbook
- Start with Agentic Audiences (UCP) if you care about agentic advertising infrastructure.
Used in AdCP Signals Adaptor + Performance & Native
- Start with the 7-Step Audit if you care about US market entry or a stalled GTM.
Used in the Market Entry Audit engagement
- Start with Behavioral Prediction if you care about psychographic activation and creative.
Used in the Mobile App Growth + Attention playbooks
- Start with the Data Collaboration Stack if you care about governed data, clean rooms, BI, and agent-ready workflows.
A seven-layer model — signal · governance · discovery · collaboration · semantic · decision · agentic. Used in the Enterprise Data Collaboration playbook.
- Start with the Agentic Operating System if you care about moving AI pilots into governed agentic workflows.
Used in the Agentic Transformation playbook
- Start with the Commercial Productization System if you care about turning capability into a repeatable product.
Used in the Commercial Productization playbook
EARO
Exposure → Attention → Relevance → Outcome. A holistic model for measuring marketing effectiveness in the age of agentic AI.
Grounded in cognitive science — specifically Dr. John Vervaeke's theories of relevance realization and collective intelligence — EARO replaces attention-only measurement with a four-component model spanning 22 measurable metrics. Explicitly models the Exposure → Attention → Emotion chain so campaigns can be evaluated on the why, not just the what.
- EExposure Was the right person reached in the right context?
- AAttention Did they notice? Did they engage?
- RRelevance Was the message meaningful and emotionally resonant?
- OOutcome Did it drive the intended business result?
Agentic Audiences (UCP)
A 77-page technical whitepaper on embeddings-based agent-to-agent advertising, covering IAB Tech Lab's Agentic Audiences — the embedding-exchange standard LiveRamp proposed as the User Context Protocol and donated in Q4 2025.
UCP defines how intelligent agents exchange compact embeddings of 256–1024 dimensions — encoding identity, contextual, and reinforcement signals in what the spec describes as a privacy-preserving, interoperable format. Fewer raw identifiers move; the controller's obligations do not lift (see the correction in Embeddings: The Next Frontier). RTB-compatible latency through binary quantization. The Vector Alignment Contract (VAC) solves the cross-model interop problem so vectors from different foundation models still produce meaningful similarity.
- Vector Alignment Contract
- Model negotiation + signed projector artifacts + graceful AdCOM fallback
- Economic Settlement
- AP2 Verifiable Digital Credentials
- Hardware-Rooted Trust
- TEE attestation, IETF RATS
- Threat Modeling
- STRIDE with pseudocode verification flows
Behavioral Prediction Framework
21 psychographic dimensions for ad activation — designed to model engagement propensity with probabilistic scoring across cognitive, emotional, and dispositional traits.
Integrates cognitive, emotional, dispositional and temperamental traits into a single model. Maps directly to adtech activation touchpoints — attention modeling, creative optimization, and media planning — with a JSON schema for real-time bidding integration.
- Cognitive Profile Fluid reasoning, working memory, verbal comprehension, processing speed
- Rationality Instrumental, epistemic, reflective, heuristic reliance
- Attachment Style Secure, avoidant, anxious, disorganized brand relationship orientation
- Big Five Personality Openness, conscientiousness, extraversion, agreeableness, neuroticism
- Temperament Profile Energy, reactivity, sociability, regulation, rhythmicity
7-Step Market Entry Audit
The operating methodology for international AdTech, MarTech, and data vendors entering the US market.
Not a whitepaper — the working framework shipped on every engagement. Time-boxed, decision-oriented. Sequence sourced from years of sell-side and buy-side practice across APAC, EMEA, LATAM, and the Americas. Each step has a measurable output.
- 01 GTM strategy development
- 02 Product-market fit assessment
- 03 Market sizing (TAM · SAM · SOM)
- 04 Sales and BD framework
- 05 Monetization and growth
- 06 Local talent and org structuring
- 07 Compliance and privacy alignment
Agentic Operating System
A six-layer model for designing agentic workflows across signals, context, tools, decisions, governance, and outcomes.
Most AI work stalls between pilot and production because the workflow has no decision rights, no governed tool access, no evaluation, and no fallback. The Agentic Operating System treats an agent as a system across six layers — so the work ships with an owner, a control plane, and a measurable business outcome instead of a demo.
- Signals What the agent senses — events, inputs, triggers
- Context Semantic infrastructure + agent-ready data
- Tools Allow-listed data + tool access, with permissions
- Decisions Decision rights + human-in-the-loop thresholds
- Governance Output policy, evaluation, monitoring, fallback logic
- Outcomes Outcome-level evaluation tied to a business metric
Commercial Productization System
A seven-layer model for turning capability into a repeatable commercial product across buyer problem, offer, pricing, proof, delivery, and expansion.
Teams that can do the work as a bespoke build often cannot sell it as a product. This system moves a capability from custom delivery to a packaged offer — buyer problem → offer architecture → packaging → pricing and value metric → proof → delivery → expansion — so each sale gets cheaper to win and easier to repeat.
- 01 Buyer problem The job the product is bought to do
- 02 Offer architecture Scope, boundaries, delivery model
- 03 Packaging Tiers, modules, editions, upgrade paths
- 04 Pricing Value metric + how price scales with value
- 05 Proof Proof taxonomy + buyer-committee narrative
- 06 Delivery POC-to-production, managed workflow, native app / API
- 07 Expansion Repeatable enterprise revenue + renewal motion
Frameworks are the start. Decisions are the point.
Use the frameworks as open-access thinking, or apply them to a live GTM, measurement, data, or operating-model problem.