Customer intent relies on mapped sources to detect and refresh intents as customers move across channels. Because each channel exposes different types of signals, coverage varies depending on whether real-time sources (Natural Language Understanding (NLU) intents, Customer Segments) or post-conversation sources (Topics) are available.

This article describes about the sources contributing to Customer intent on each channel so you can understand where detection is automatic, where it is delayed until after the conversation, and where no native signals exist today. This understanding helps you to plan your taxonomy, ensure that each intent has at least one reliable signal, and set realistic expectations for detection across your digital and voice journeys.

Definitions

• Category — High-level grouping of customer intents. For example, claims, lending, account management.
• Customer intent — The reason for engagement. Customer intents are associated to an External Contact, persists and expires when the intent is inactive for a contact beyond a defined expiration period. By default, the expiration day is set to seven.
• Sources (signals) — What assigns or refreshes a customer intent for the contact.
  • NLU intents (Native Genesys Bots, Virtual Agents, and Copilots)
  • Customer Segments (web or mobile app behavior)
  • Topics (Speech and Text Analytics, post‑conversation)