A public or external claim should point back to an identifiable source, publisher, observation date, or retrieval date whenever practical.
Portfolio research standards / 2026
Evidence before claims.
ExpenseIntel, Waterline Intel, and Gage Grid work in different markets, but they should follow the same evidence discipline: know what is observed, what is modeled, what is still unknown, and what has actually been verified.
A nearby asset, market observation, public listing, benchmark, or model input is context until the decision-specific fact is confirmed.
Products preserve when a source was observed separately from when Queenan Capital retrieved it so stale data cannot appear current.
Modeled estimates can guide screening and stress tests, but they do not become quotes, utility commitments, capacity confirmations, or executed outcomes.
A written quote, utility response, public record, or user-confirmed outcome should replace weaker modeled assumptions rather than sit beside them as an equal fact.
Contradictory, stale, rejected, or failed evidence should remain visible when it materially changes a decision.
The strongest future moat is not more features. It is the growing record of what was quoted, verified, executed, paid, delivered, and changed.
EVIDENCE CLASSES
User. Public. Verified. Modeled. Unknown.
The goal is not to eliminate modeling. The goal is to make the hierarchy legible. Direct, written and outcome evidence should outrank generalized public context; public context should outrank unsupported assumptions; and unresolved facts should remain unknown until they are replaced.