What exists
The Delphic Question-Family Demand Estimator is a protected method for estimating demand at the same granular unit used to map knowledge, measure AI position and build an Answer Object.
Its unit is one validated question family: several natural buyer questions that require substantially the same answer. The estimator does not treat a broad website topic, a keyword cluster or total site traffic as an equivalent market.
| Record field | Public description |
|---|---|
| Owner | Delphic |
| Unit measured | One validated question family |
| Primary evidence class | Traffic and live behaviour attached to the relevant Knowledge Index objects |
| Secondary evidence class | Selected external signals that can provide context where direct evidence is limited |
| Output | A dated estimate of query volume and direction, with a confidence label and interpretation boundary |
| Intended use | Comparing demand across mapped question markets and informing where investigation or investment may be justified |
This record confirms that the estimator exists and what it returns. It is not the estimator itself.
What this record establishes
AI platforms do not provide a complete public counter for every commercially meaningful question buyers may ask. Delphic therefore treats demand as an estimate rather than a directly observed universal total.
Live Knowledge Index behaviour is valuable because each relevant public object is attached to a defined question family. Activity can therefore be interpreted at the level of a knowledge market rather than a broad website category. Selected proxy evidence can add context, particularly before enough direct behaviour has accumulated.
The estimator returns a decision input, not a promise. A low observation may reflect weak demand, weak discovery, incomplete mapping, limited evidence or a short observation window. The confidence and interpretation boundary supplied with the estimate are part of the result.
The broader role of this demand record is explained in question-market intelligence. That framework shows how estimated demand joins AI position, public-supply movement and treatment evidence. This protected record owns only the conversion of demand evidence into a per-family estimate.
Access and verification
Application begins with a validated question family and the evidence that is legitimately available for it. Delphic returns the estimate with its date, evidence window, confidence and limits rather than presenting a bare number.
The client can inspect the public and first-party evidence made available to the engagement, compare the estimate with later behaviour and measure independently. Independent checking is compatible with the method remaining protected: the result can be assessed without publishing the rules that produced it.
The output can then inform the protected question-market opportunity ranking or remain an observation where the evidence is not strong enough to support a priority.
What remains protected
The public record does not identify the complete signal set, inclusion tests, attribution rules, transformations, relative importance of inputs, adjustments for missing or uneven evidence, confidence construction or checking thresholds.
Those elements are the substance of the estimator. Publishing them would turn a source record for protected knowledge into a reproduction guide. Delphic instead makes the estimator's ownership, scope, output, limits and access path machine-legible while keeping the operating method private.