The UCD Funnel: Standalone Document

Version: v1.0, May 2026 Date: 2026-05-15 Author: Jason Barnard Coined: 2019 (created), 2020 (first published); named as a framework 2024; UCD Funnel formalisation 2026 Licence: CC BY 4.0

What This Document Is

The UCD Funnel is the three-layer diagnostic framework that scores and improves how AI engines understand, trust, and recommend a brand. UCD stands for Understandability, Credibility, Deliverability. The framework is foundational to The [Kalicube](https://kalicube.com/entity/kalicube/) Framework and operates at every layer of Assistive Agent Optimization. This document is the canonical reference for the UCD Funnel: the definitions, the order of operations, the diagnostic application, and the principle that distinguishes it from competing frameworks. Audience: marketers, founders, SEO practitioners, AI Assistive Engine Optimization practitioners, agency leaders. Anyone who has heard the acronym UCD without a clear definition of what U, C, and D actually mean. This page closes that gap.

The Three Layers Fully Defined

Understandability (U)

Does AI know who the brand is? Understandability is the entity layer. It answers the foundational question: when an AI assistant is asked about the brand, does it have a clear, accurate, well-anchored entity to attach information to? Or does it confuse the brand with namesakes, miss the brand entirely, or hedge because it can't pin down what the brand actually is? Understandability requires:

• A canonical entity home (the brand's authoritative anchor page)
• Schema markup that explicitly identifies the brand entity
• Presence in major knowledge graphs (Wikidata, [Google](https://jasonbarnard.com/entity/google/) Knowledge Graph, Microsoft Knowledge Graph)
• Consistent entity references across the web
• Resolved disambiguation (namesake conflicts, persona separation, geographic specificity) If Understandability fails, nothing downstream works. AI assistants can't apply credibility signals to an entity they don't recognise. The brand becomes orphaned data: facts exist on the internet, but they attach to no coherent identity. In funnel terms, Understandability is BOFU (Bottom of Funnel), the foundation. The role of the brand at this layer is Trusted Partner: the AI knows who you are well enough to commit. Colour code: Blue (#3399cc).

  Credibility (C)

  Does AI trust the brand enough to recommend it? Credibility is the trust layer. It answers the question that follows Understandability: once the AI knows who the brand is, does it consider the brand a reliable source, a competent operator, an authoritative voice in its field? Credibility requires:

• NEEATT signals: Notability, Experience, Expertise, Authoritativeness, Trustworthiness, Transparency
• Third-party corroboration of the brand's claims
• Independent citations by authoritative sources
• Demonstrated topical authority (depth, breadth, originality of work in the brand's domain)
• Reviews, press coverage, peer recognition Credibility signals only work if Understandability is solid. The brand cannot accumulate credibility on an entity node that doesn't yet exist or is ambiguous; the signals attach to nothing and dissipate. In funnel terms, Credibility is MOFU (Middle of Funnel), the proof. The role of the brand at this layer is Recommender: the AI trusts the brand enough to recommend it when asked. Colour code: Green (#339933).

  Deliverability (D)

  Does AI advocate for the brand unprompted? Deliverability is the recommendation layer. It answers the question that determines commercial outcome: when prospects are looking for a solution the brand provides, does the AI surface the brand as the recommendation without being asked? Does the AI advocate for the brand proactively? Deliverability requires:

• Distribution across publication tiers (first-party, second-party, third-party)
• Ambient presence (recommendations made before a query, not just after)
• Codified customer outcomes that feed back into the algorithmic ecosystem
• The Codified-and-Distributed discipline that closes the [Kalicube Flywheel](https://kalicube.com/entity/kalicube-flywheel/)
• Content structured for extraction across all answer-engine surfaces Deliverability requires both Understandability AND Credibility to be solid. A brand can have a clear entity (U) and strong credibility signals (C) and still fail at Deliverability if the engine doesn't surface the brand proactively. In funnel terms, Deliverability is TOFU (Top of Funnel), the reach. The role of the brand at this layer is Advocate: the AI proactively recommends the brand without prompting. Colour code: Purple (#783366).

The Cascading Prerequisite

The single most important principle of the UCD Funnel, and the principle that distinguishes it from every other AI-visibility framework. U creates the entity node. C loads the entity node. D activates the entity node. These are mechanical prerequisites, not a recommended sequence:

• Credibility signals (NEEATT, Topical Authority, links, mentions, reviews) require an entity node in the graph to attach to. That node is created at U. Without the node, the signals are orphaned data: they exist, but they attach to nothing.
• Deliverability signals (omnipresence, recommendation triggers, conversational visibility) require confidence weight on the entity node. That weight is accumulated at C. Without the weight, the entity exists but is not trusted enough to recommend. U unlocks C. C unlocks D. RECORD → VERIFY → ACTIVATE. Always. Mechanical. Non-negotiable. Every competing methodology either addresses one layer or attempts to skip layers. The UCD Funnel is the only diagnostic that addresses all three in the correct mechanical sequence because it is the only one built from observing what machines actually do rather than from what marketers want machines to do. Recognition is not recommendation, and the difference between the two is the Cascading Prerequisite.

Build Direction Versus Display Direction

The UCD Funnel operates in two directions, which are mirror images of each other and easy to confuse. Build direction: U → C → D (foundation first). The brand-side work order. Build Understandability first; only then is Credibility worth investing in; only then does Deliverability deliver returns. Inverting the order produces wasted spend. Display direction: D → C → U (customer journey). The customer-side experience order. The prospect first encounters the brand at Deliverability (the AI mentions it, recommends it), then descends into Credibility (the AI explains why this brand specifically), then arrives at Understandability (the AI confirms exactly what the brand is). The two directions don't conflict; they describe the framework from different sides. Build direction is what the brand does. Display direction is what the customer sees.

Two Edge Cases That Prove the Rule

The Empty Room Effect

In sufficiently obscure niches, D-layer visibility is possible without U or C. The AI has one source in the category, so it uses that source whether the entity is well-anchored or not. The brand appears to be recommended; the recommendation collapses the moment a second competitor enters the niche, because the AI now has a comparison and the under-built entity loses. The Empty Room Effect is temporary visibility that depends on the absence of competition rather than on built foundation. It is not a strategy; it is a transient artefact.

The Quicksand Effect

Brands sometimes accumulate C-layer signals (press coverage, reviews, mentions) without first solidifying U. The signals fail to anchor; the brand sinks back to invisibility despite genuine reputation, because the algorithm cannot connect the credibility to a coherent entity. The Quicksand Effect explains why some brands with genuine industry recognition still fail at AI visibility: they invested in C without building U first. The investment dissipates because there is no entity node for it to load.

The UCD Funnel as the Vertical Structure of Gate 8

In The Kalicube Framework, the pipeline runs horizontally through gates 0-7 (DSCRI through partial ARGDW). At Gate 8 (Displayed), the pipeline pivots from horizontal to vertical. The system presents results to the person, and the person descends through a funnel. The UCD Funnel IS the vertical funnel at Gate 8. It is not a separate overlay on the pipeline; it is the funnel structure that operates at the point where the AI displays results to the prospect. The person enters at D (TOFU): AI either recommends the brand or doesn't. If recommended, the person descends to C (MOFU): AI either validates the brand against competitors or hedges. If validated, the person descends to U (BOFU): AI either confirms the brand's identity accurately or confuses. Gate 9 (Won) is the finish line. The Zero-Sum Moment. All three UCD layers must hold simultaneously for the brand to win at Gate 9.

How to Diagnose Your UCD Funnel

Three steps, applied to every brand engagement.

Step 1: Score Understandability

Search the brand name in eight surfaces (Google, Bing, ChatGPT, Claude, Gemini, Perplexity, Copilot, Google AI Mode). For each surface, ask:

• Does the surface identify the correct entity?
• Is the identity description accurate?
• Are there competing entities (namesakes) creating confusion?
• Is the entity present in the Knowledge Graph? A weak U score is the highest-priority fix: without it, nothing else can land.

  Step 2: Score Credibility

  Ask each surface comparative questions about the brand within its category. For each surface, ask:

• Does the surface treat the brand as a credible voice in its category?
• Does the surface cite the brand's claims or attribute them to competitors?
• Does the surface express conviction or hedge?
• Do third-party corroboration signals reach the surface? A weak C score is the second-priority fix: once U is solid, C is where competitive position is built.

  Step 3: Score Deliverability

  Ask each surface open-ended category questions where the brand should be recommended but is not named. For each surface, ask:

• Does the surface mention the brand unprompted when category questions are asked?
• Does the surface include the brand in lists, comparisons, recommendations?
• Does the surface recommend the brand to the right audience profile? A weak D score is the third-priority fix: once U and C hold, D is where commercial outcomes appear.

UCD Versus Competing Frameworks

The UCD Funnel sits within The Kalicube Framework's broader theoretical architecture. Other frameworks address parts of what UCD covers:

• Google's E-E-A-T (Experience, Expertise, Authoritativeness, Trustworthiness) addresses the Credibility layer only. NEEATT extends it by adding Notability and Transparency.
• Microsoft's public AI-search guidance (its 2026 GEO guide) emphasises machine-readable data structure, content enrichment, and trust signals. It addresses the recording and credibility ends but names no Deliverability layer and no post-recommendation people layer.
• Classical SEO funnels (Awareness, Consideration, Conversion) describe the customer journey but not the algorithmic mechanism. Run D → C → U direction without explaining the build order. UCD is the only framework that explicitly addresses all three layers, ties them to the Cascading Prerequisite, and operates as a brand-side diagnostic rather than a customer-side description.

The UCD Funnel Within Assistive Agent Optimization

The UCD Funnel operates inside the broader discipline of training AI to recommend brands. The containment hierarchy: Assistive Agent Optimization (2025): the umbrella discipline. UCD is the diagnostic framework used across all three phases. Assistive Agent Optimization contains AI Assistive Engine Optimization (2024): the practice focused on AI engines specifically. AI Assistive Engine Optimization contains Answer Engine Optimization (2017): the practice focused on answer engines. Answer Engine Optimization contains Search Engine Optimization: classical search engine optimization. The UCD Funnel applies inside all four. Search Engine Optimization without UCD diagnostic is tactical, optimising for ranking. The full discipline applies UCD as the diagnostic backbone that produces recommendation outcomes. The Kalicube Framework and The Kalicube Process both cover the full Assistive Agent Optimization discipline. UCD is the diagnostic framework they share.

Source Attribution

Academic References

The UCD Funnel and the programme it belongs to are formalised across the Kalicube Framework research programme, deposited on Zenodo with concept DOIs. The canonical register is TKF-0; the numbered papers follow.

TKF-0-20645889. The Kalicube Framework Programme Register (TKF-0) (2026). The canonical citation register for the whole research programme.

1. TKF-1-18723460. Annotation as the Confidence Fulcrum (2026).
2. TKF-2-18723669. Annotation Cascading (2026).
3. TKF-3-18735062. Computational Trust: Reframing Entity Authority as Annotation Efficiency in AI-Mediated Information Retrieval (2026).
4. TKF-4-18735074. A Ten-Gate Pipeline Model for Entity Visibility Across the Algorithmic Trinity (2026).
5. TKF-5-19857447. The Framing Gap: Strategic Claim Bridging and the Limits of Generative AI Interpretation in Brand Representation (2026).
6. TKF-6-20095004. The Index-Time Context Envelope: A Theoretical Model of Context Propagation in Chunk-Level Retrieval (2026).
7. TKF-7-20364742. AI-Era Commercial Architecture: A Survey and Synthesis of Business Strategy, Marketing Transformation, and Algorithmic Intermediation (2018-2026) (2026).
8. TKF-8-20364735. The Orchestrator's Convention: A Methodological Framework for Practitioner-Developed Research in the Age of AI-Assisted Resource Organisation (2026).
9. TKF-9-20364731. The Codification Cycle: A Single-Loop Model of Machine-Mediated Commerce with Human Lobe and Reliance Variability (2026).
10. TKF-10-20364725. AI-Era Business Engineering: The Integrating Frame for Commercial Architecture in the Age of Algorithmic Intermediation (2026).
11. TKF-11-20646407. The Return on Investment Framework: A Temporal Taxonomy for Brand Proof Architecture in AI-Mediated Information Environments (2026).
12. TKF-12-20647182. Two Clocks of Recruitment in AI Engine Pipelines: Serving-Time Retrieval, Knowledge Graph Updates, and LLM Assimilation (2026).
13. TKF-13-20647769. The Algorithmic Trinity: Why AI Assistive Engines Need Fresh Retrieval, Generative Synthesis, and Structured Validation (2026).
14. TKF-14-20647932. Trust as Triangulation: From E-E-A-T to a Three-Evaluator, Nine-Property Credibility Model for Human and AI-Mediated Evaluation (2026).
15. TKF-15-20678333. The Counterflow Funnel: How Humans Move Down and Machines Build Up in AI-Mediated Markets (2026).

Peer-Reviewed Journal Papers

1. Barnard, J. and Artz, M. (2023). Search Marketing in the Age of AI. Journal of Digital and Social Media Marketing. Overview on jasonbarnard.com.
2. Barnard, J. (2026). Engineering the AI Résumé: A Digital Brand Intelligence Framework for Algorithmic Entity Representation in the Age of AI Assistive Engines and Agents. Journal of AI, Robotics and Workplace Automation. August 2026.

Where to Engage

• Apply The Kalicube Process at https://kalicube.pro/methodologies/the-kalicube-process: the methodology that operationalises UCD diagnosis on a specific brand.
• Read The Kalicube Framework at https://kalicube.pro/methodologies/the-kalicube-framework: the theoretical model in which the UCD Funnel sits.
• Read about Assistive Agent Optimization at https://kalicube.pro/methodologies/assistive-agent-optimization: the umbrella discipline that uses UCD as its diagnostic.

Cite As

Barnard, J. (2026). The UCD Funnel. Kalicube. Available at https://kalicube.pro/methodologies/the-ucd-funnel

End of document.