prescore

Prescore is a term used across multiple industries to denote an initial or provisional numerical assessment assigned before a final scoring or decision stage. Unlike a finalized score, a prescore is typically based on partial information, a subset of features, or an interim model, and is intended to guide subsequent actions rather than serve as a final verdict.

In lending and credit workflows, prescore is often used to quickly screen applications, triage risk, or determine eligibility for further underwriting. It is usually derived from readily available data such as basic consumer information, transaction history, or historical behavior, and is subsequently refined by a full underwriting or risk score when more data becomes available. In admissions and hiring processes, prescore may provide an early estimate of a candidate's potential outcome to prioritize interviews or further assessment.

In data science and analytics, prescore can be a preliminary model output that serves as an input to a final scoring stage or as a feature in an ensemble. It may be updated as new data arrives and can help reduce latency by providing quick recommendations or routing decisions. In marketing, advertising, or search, prescores can estimate relevance or expected performance before more expensive evaluation steps are performed.

Because the meaning and computation of prescores vary by context, organizations should document the methodology and calibrate prescores against actual outcomes. Misuse can introduce bias or misallocation of resources, so prescores are generally treated as interim indicators rather than final judgments. See also scoring model, predictive analytics, risk assessment.