Booking intelligenceIndia · United States

Editorial and data methodology

What the number knows—and what it does not.

Every public report starts from a saved provider reading. We preserve measurement, carried state and uncertainty as different claims.

Measured

Read directly from the provider during the observation shown on the report.

Carried

A prior exact reading retained because the provider no longer lists the show or a safe refresh was unavailable.

Unknown

Inventory we cannot honestly classify. It is never silently converted into a sale.

Sources

India reports currently use District and BookMyShow. USA reports use Fandango. GSC Malaysia and Shaw Singapore remain limited provider pilots and are not part of the public scheduled pipeline.

Saved evidence first

The collection stage writes the provider response into a durable local run with coverage, errors and timestamps. Public publishing reads that saved run; it does not ask a live provider again.

No double counting

Where District and BookMyShow cover the same Indian theatre and show, the merge layer uses conservative theatre identity and show multiplicity checks before counting it once.

Approval is the authority

Deterministic code builds the figures, card and public body. Telegram shows the exact artifact with Approve, Reject, Edit text and Ask AI to change. AI may reword supported text but cannot introduce a new number. Only a configured human approver can publish.

Immutable reports

A published report cannot be silently rewritten through the publication API. A correction or later observation receives a new report identifier and publication timestamp.

Not box-office reporting

Advance-booking readings are not final theatrical gross, distributor share, footfall or a forecast. Provider labels can be categorical, and blocked or held seats are not automatically bookings.