A stockout alert fires on a Tuesday. It sits in a dashboard nobody opened until Thursday. By then the order already missed the ship window. The alert worked exactly as designed. The gap wasn't detection. It was everything after.
Definition
The condition is detected. Stock cover falls under threshold, a PO passes its promise date, a machine logs an anomaly, an invoice fails three-way match. This is the layer the market has already solved.
Most mid-market operations already have Signal. Demand forecasting, anomaly detection, and inventory risk scoring fire the alert on time, every time. The five-layer model treats Signal as the stage that's already working in most operations. That's why the Decision Latency Diagnostic scores it separately from the four that follow: a strong Signal score next to weak Route or Approve scores is the most common pattern IntelliConnectQ sees in Business Central environments.
Why does Signal usually work while the other four stages don't?
Signal is a detection problem, and detection is what forecasting, BI, and anomaly-scoring tools were built to solve. Route, Approve, Execute, and Audit are ownership and accountability problems: who acts, at what threshold, with what record. Most operations never built anything to answer those questions. That's why a team can have a genuinely strong Signal layer and still take two to three days to act on what it detects.