Resource center
Specsheets
Glossary
Conversational Commerce
Buying and selling through real-time dialogue, whether with an AI agent, a live associate, or a combination of both. The transaction happens inside the conversation, not alongside it.
AI Shopping Agent
A software agent embedded in an online store that guides customers through product discovery, comparison, and purchase using natural language. Not a support chatbot. Designed to sell.
Product Discovery
The process by which a customer finds the right product. In traditional ecommerce, this relies on search bars, filters, and category pages. AI-driven product discovery uses behavioral signals and natural language to surface relevant products proactively.
Behavioral Fingerprint
A real-time profile built from device data, navigation patterns, and interaction signals. Used to understand anonymous visitors from the first pageview, before any account or purchase history exists.
Recommendation Engine
A system that suggests products to a customer based on data inputs like purchase history, browsing behavior, and contextual signals. Ranges from simple rule-based logic ("customers also bought") to real-time AI models that reason across multiple data sources.
Intent Recognition
The ability to interpret what a customer is trying to do based on their actions, not just their words. Distinguishes between browsing, comparing, and buying behavior.
Catalog Intelligence
Structuring and indexing a product catalog so that an AI system can query it in natural language, understanding attributes, variants, compatibility, and availability.
Knowledge Grounding
Connecting an AI agent to verified, store-specific information (return policies, shipping terms, product specs) so that responses are accurate rather than generated from general training data.
Omnichannel
A unified customer experience across multiple touchpoints: web chat, WhatsApp, Instagram, Messenger, email. One agent, one configuration, every channel.
Human Handoff
The process of transferring a conversation from an AI agent to a live human associate, with full conversation context preserved so the customer does not have to repeat themselves.
Cart Abandonment
When a customer adds products to their cart but leaves without completing the purchase. AI-driven interventions (personalized nudges, better recommendations, conversational re-engagement) can reduce abandonment rates significantly.
Zero-Party Data
Information a customer intentionally shares: preferences, quiz answers, stated needs. Distinct from behavioral data, which is observed. Both feed into AI decision-making.
First-Party Data
Information collected directly from your customers through their interactions with your store: purchases, browsing, account details, support conversations. The foundation of any serious personalization effort.
Average Order Value (AOV)
The mean amount spent per order. A primary lever for revenue growth, often raised by smarter recommendations, complementary bundling, and conversational guidance toward higher-value choices.
Customer Lifetime Value (CLV)
The total revenue a customer is expected to generate across the entire relationship with your store. AI-driven retention and personalization aim to grow CLV rather than chase one-off conversions.
Conversion Rate
The share of visitors who complete a desired action, typically a purchase. A useful metric, but a thin one. It does not capture what happened in the visits that did not convert, which is where intent signals live.
Customer Journey
The end-to-end path a customer follows from first contact to purchase and beyond. Rarely linear in practice. AI systems read journeys as networks of signals, not funnels.
Personalization Engine
A system that adapts what a customer sees in real time, based on who they are, what they have done, and what they appear to want. Goes beyond "recommended for you" to shape the entire shopping experience.
Large Language Model (LLM)
A neural network trained on large volumes of text, able to understand and generate natural language. The underlying engine that lets AI shopping agents converse fluently. Without grounding, an LLM is an unreliable narrator.
Retrieval-Augmented Generation (RAG)
A pattern where an AI model retrieves relevant store-specific information at query time and uses it to compose accurate answers. The standard approach to keeping AI agents grounded in your real catalog and policies.
Semantic Search
Search that understands meaning rather than matching keywords. A query for "something warm for hiking" returns insulated jackets, not just products containing the word "warm." Powered by vector embeddings of your catalog.
Customer Data Platform (CDP)
A system that unifies customer data from every source (web, mobile, support, CRM) into a single profile. The clean foundation an AI engine needs to make trustworthy real-time decisions.
Cross-Sell
Suggesting complementary products to what a customer is already considering. Done well, it feels like advice. Done badly, it feels like a banner ad.
Upsell
Guiding a customer toward a higher-value option than the one they were originally considering, when it genuinely fits their needs. The line between upsell and pushiness is drawn by relevance.
Headless Commerce
A store architecture where the frontend is decoupled from the commerce backend, communicating via APIs. Makes it easier to integrate AI components, custom storefronts, and new channels without rebuilding the platform.