Coveo rolls out Coveo Conversational Product Discovery

Executive Summary

• Coveo Solutions Inc. has launched "Coveo Conversational Product Discovery," a new capability within its Coveo for Commerce platform.
• The feature integrates natural-language conversational AI directly into the core search experience, allowing shoppers to discover products through dialogue rather than traditional keyword searches.
• The solution is designed to replace fragmented standalone chat bots by maintaining search speed, relevance, and merchandising control while enabling context-aware product refinement.

Key Details

• Product Name: Coveo Conversational Product Discovery.
• Platform: Available as an add-on feature for Coveo for Commerce.
• Core Functionality:
  • Allows shoppers to describe needs using natural language.
  • Provides curated product results grounded in catalogue data.
  • Enables refinement of results through follow-up questions.
  • Supports product comparison and key attribute analysis.
  • Facilitates the building of bundles for complementary products.
• Technical Architecture: Built on Coveo's agentic orchestration architecture, which coordinates multiple AI capabilities to interpret intent, retrieve products, and assemble responses based on catalogue data and merchandising rules.
• Control Mechanisms: Retail teams retain control via defined layouts, content guardrails, and merchandising directives to shape result appearance.
• Key Benefits:
  • Converts exploratory traffic into revenue by guiding vague needs to curated selections.
  • Augments existing high-performing search experiences rather than replacing them with chat bots.
  • Maintains merchandising control through deterministic layouts and policy-driven AI execution.
  • Reduces dead-end experiences via guided refinements and next-best suggestions.
• Availability: Available today.

Notable Quotes

• "Shoppers don't think in keywords," said Peter Curran, general manager of commerce at Coveo. "Imagine if someone walked into a store and just said 'shoes.' Unfortunately, most digital stores can handle that only if they're lucky."