How Value Actually Flows (And Why Palantir's Entity Model Misses It)

Palantir's ontology is built on entity-relationship modeling. Object types, properties, link types. Map the nouns and the connections between them. This is powerful — it turns scattered databases into a navigable model of reality.

But entity-relationship models have a blind spot. They map what exists. They don't map how value moves.

Static vs. Dynamic

"Supplier A connects to Manufacturer B" is an entity-relationship statement. It tells you the structure exists. It doesn't tell you that the handoff between them costs $2 million in delays per quarter. It doesn't tell you that value is leaking at that junction because the information flow is asymmetric. It doesn't tell you that three other organizations have solved this exact bottleneck in different ways.

Entity models describe structure. Value flow analysis describes dynamics. Both matter, but the actionable intelligence — the kind that changes decisions — almost always lives in the dynamics.

REA Semantics

We model value flow using REA — Resource-Event-Agent — the accounting ontology that was designed to capture economic exchange, not just entity relationships.

Resources aren't just labeled. They're classified across 16 dimensions: what the resource is, what it can do, who has rights to it, what constraints apply, where it's stuck, and what would unlock it. A "partnership opportunity" isn't just a node in a graph. It has a rights structure, a constraint profile, a flow state, and a score indicating how much friction exists between its current state and its optimal state.

Events capture the verbs — 23 value flow action types that map the movements of business. Not "Supplier A connects to Manufacturer B" but "Supplier A transferred inventory to Manufacturer B on March 3rd, triggering a quality review event that created a 6-day delay."

Agents are the actors — people, organizations, AI systems, or automated processes that participate in value exchanges. Each agent has preferences, constraints, and a relationship to the resources they hold or need.

Scored Opportunity Maps

When you model an organization's ecosystem with REA semantics instead of flat entity-relationships, you get something Palantir's model doesn't produce: scored opportunity maps.

Every value flow gets scored for friction, blockage, and leakage. The bottleneck map emerges — not as a hypothesis, but as a measured property of the system.

We overlay this on a Business Model Canvas to create strategic visualizations. Revenue streams, cost structures, key partnerships, and customer segments — all annotated with flow scores that show where value moves freely and where it gets stuck.

The pattern we see consistently: organizations sitting on value they've already created but can't access. The connections exist. The resources exist. The flow is blocked at specific, identifiable points. Once you map the flow, the interventions become obvious.

Entity models tell you what's there. Value flow tells you what to do about it.