AI-enabled lead temperature: a narrative review and five-dimension framework for machine-learning lead prioritization

Purpose This study aims to reconceptualize cold, warm, hot and sales-ready leads as dynamic states of sales-readiness and develops a five-dimension framework based on customer-profile fit, engagement, buying intent, authority and timing. Design/methodology/approach A narrative review integrates the original 47-source DOI-bearing peer-reviewed analytical corpus with a reviewer-directed verification and update of recent peer-reviewed evidence. The synthesis covers qualification/scoring, nurturing, marketing-sales handoff, CRM/pipeline governance, inside sales/follow-up and AI/predictive lead management, using explicit eligibility, coding and bias-mitigation procedures. Findings Lead temperature is a reversible, timestamped configuration rather than a fixed label or engagement score. The framework translates the five dimensions into four operational levels: cold, warm, hot and sales-ready. Machine-learning models can estimate or rank this state, but decision value depends on recency, calibration, explainability, data quality and managerial governance. Research limitations/implications The framework requires longitudinal, multi-firm validation across channels, industries and sales-cycle lengths. Practical implications RevOps, marketing and sales teams can use the model for routing, nurturing, response-time SLAs, downgrading stale leads and governing predictive scores. Originality/value The paper links practitioner lead-temperature terminology to machine-learning lead prioritization through an interpretable feature architecture and eight testable propositions.