Risk often emerges along predictable stages of the buyer lifecycle. This chapter models buyer-side abuse across signup, authentication, browsing, incentive usage, checkout, and post-purchase interactions, emphasizing that signals are stage-specific and temporal. Early stages focus on identity quality, device continuity, and velocity patterns that indicate automation or multi-accounting. Mid-journey behavior captures scraping, narrow repetitive navigation, funnel manipulation, and voucher exploitation, where intent may be inferred before loss materializes. Later stages address refunds, returns, and dispute behavior, where outcomes are delayed and partially observed. The chapter stresses proportional intervention: lightweight friction early, stronger controls as risk accumulates, and escalation paths that remain measurable and explainable. By treating events as sequences rather than independent points, buyer-journey models use continuity signals to improve accuracy and reduce false positives. Design guidance includes how to structure stage-specific labels, how to prevent leakage across steps, and how to coordinate interventions so that friction is applied only when it changes expected loss. Examples focus on aligning model outputs to controls such as step-up checks, throttling, and review queues, with explicit friction budgets. The result is a practical blueprint for linking behavioral telemetry to stage-specific objectives while balancing loss prevention with conversion and customer experience.