Methodology

The existing publication review separates novelty, validation, commercialization evidence, company evidence, reader value, and source clarity to organize public and retained records.

The Commercialization Watch publication-time signal score is not a success probability or investment opinion. It prioritizes follow-up using only evidence available when the paper was published, such as prototypes, manufacturing, reliability, or clinical validation.

Later patents, company activity, clinical progress, and production events never overwrite the original score. Verified later events are recorded separately as commercialization momentum and timeline entries. Domain-specific rules recognize that wafer and yield evidence matters for semiconductors, cycle life for batteries, and clinical evidence for biotech.

A signal is not stored without a supporting excerpt and source URL. A patent that has not been checked is shown as not checked, not as no patent. Missing evidence and caution signals remain visible.