E-commerce Chat Glossary
Plain-language definitions of the terminology we use across Emporiqa docs, blog posts, and marketing pages. Each entry has a stable anchor so you can link directly to it.
Index
- AI chatbot #
-
An AI-driven chat interface on a website that talks to users in natural language. On an e-commerce store, an AI chatbot answers shopper questions, recommends products, and can walk the shopper to checkout.
AI chatbot is the category label shoppers and store owners type into search engines. The term covers everything from simple FAQ bots to fully conversational agents that read catalog data and close sales. Emporiqa is an AI chatbot for e-commerce stores; inside the category, it behaves like an online salesperson (recommends a specific product, handles objections, suggests variations, walks the shopper to checkout) rather than deflecting to a lead form. The terms chat assistant, AI chatbot, and online salesperson describe the same product from three angles: admin-UI tool, SEO category, and shopper-facing behavior.
See also: Features, Integrations
- Chat assistant #
-
A conversational interface on an e-commerce store that answers shopper questions from the store's catalog and policies. Same product as an AI chatbot; the term is used in admin UI and developer contexts.
A chat assistant differs from a generic chatbot in one critical way: it reads structured catalog data (products, attributes, variants, pages, policies) rather than scraping storefront HTML. The result is variant-aware answers, accurate stock and price data, and no breakage on theme updates. The best chat assistants go beyond answering: they act as an online salesperson, read shopper intent, recommend a specific product with reasoning, handle objections (price, fit, durability), suggest the right variation, and walk the shopper to checkout. Emporiqa is built to behave this way by default.
See also: Integrations
- Online salesperson #
-
The behavior frame for Emporiqa: an AI chatbot that sells. Reads shopper intent, recommends a specific product, handles objections, and walks shoppers to checkout rather than deflecting.
Online salesperson describes what the chatbot does, not a new product category. Where a typical chatbot deflects ("Let me know if you need help"), an online salesperson engages like a human one would: qualifies the shopper ("gift, upgrade, or something specific?"), recommends one product with reasoning tied to their need, handles objections ("it's pricey" / "is it durable?") with concrete product data, suggests the right variation ("128GB Midnight, plenty of storage, easy to spot"), and closes by adding to cart. Emporiqa's agents are tuned to this behavior by default, in 65+ languages, grounded in the store's own catalog. Visual search is included: shoppers upload a photo, the chat returns the matching or closest item from the catalog.
See also: Features, Integrations
- Chatbot vs chat assistant #
-
Same product, two names. Shoppers and search engines use "AI chatbot". Store admin UI and developer docs use "chat assistant". Both refer to the conversational layer you add to your store.
The two terms are synonymous in practice and Emporiqa uses both depending on audience. "AI chatbot" is the SEO category (what buyers search for) and the primary claim slot. "Chat assistant" is the name inside the admin dashboard, in integration docs, and in developer-facing content where a persona noun would feel off. When an article compares "chatbots" to "chat assistants" as if they were different products, the distinction is usually marketing positioning rather than a real architectural difference. What matters is what the chat does: grounded answers from your catalog, objection handling, variation suggestions, add-to-cart, and handoff.
See also: Features
- E-commerce chatbot #
-
An AI chatbot purpose-built for online stores. Reads the store's catalog and policies, recommends products, answers shipping and returns questions, and can add items to cart.
A generic chatbot answers questions from whatever content you feed it. An e-commerce chatbot is specialized: it understands variants (size, color, SKU), tracks price and stock in real time, respects per-channel or per-locale catalog scoping, and can interact with the store's cart and checkout. That specialization is what separates a generic LLM wrapper from a tool that drives revenue. Emporiqa is designed end-to-end for e-commerce: native plugins for WooCommerce, Magento, PrestaShop, Shopware, Drupal Commerce, and Sylius feed structured catalog data through webhooks, and the chat uses that data to answer and sell.
See also: Integrations
- Conversational commerce #
-
Selling through conversation rather than through static category pages and search result lists. The shopper describes what they want; the store responds with a product, price, and an add-to-cart action.
Conversational commerce covers any selling channel where the interface is a chat: AI chatbots on the store, messaging apps (WhatsApp, Messenger), voice assistants. The common thread is that the shopper expresses intent in natural language and the store responds in kind. On an e-commerce site, this shortens the path from question to purchase: "warm jacket under 100 euros in size M" becomes a product card with the right size already selected, instead of ten minutes navigating filters. Emporiqa delivers conversational commerce inside the store itself, no third-party platform required.
See also: Features
- Product recommendation AI #
-
Software that suggests relevant products to a shopper based on their intent, context, or history. In a chatbot context, the AI picks one or a small number of products that match the shopper's stated need.
Traditional recommendation engines use purchase history and collaborative filtering to suggest "people also bought" rows. An AI chatbot operates in a tighter loop: the shopper states a need in the current conversation, and the chat returns one product with reasoning tied to that need. Emporiqa uses hybrid search (semantic + keyword) to retrieve candidates from your catalog, then the chat presents the best match with price, stock, and an add-to-cart button. Recommendations are grounded in your own catalog, not a generic product graph, so everything the shopper sees is in stock and linked to your checkout.
See also: Hybrid search
- Hybrid search #
-
A product search strategy that combines semantic (vector) search with keyword (BM25) search and fuses the results.
Vector search finds items by meaning: a query for 'warm jacket' surfaces insulated parkas even if the product page never uses the word 'warm'. BM25 handles exact matches like SKUs, brand names, and model numbers. Hybrid search runs both in parallel and merges results using reciprocal rank fusion, so you get the best of each.
See also: How product search works
- Vector search #
-
Searching by semantic meaning using embeddings: numerical representations of text that capture meaning.
Products, questions, and policies are converted to high-dimensional vectors by an embedding model. A query is converted the same way. The search finds the closest vectors in space. This is how 'gift for dad' can return tools, grilling gear, or headphones without any keyword overlap.
- BM25 #
-
A classical keyword-based ranking algorithm. Scores documents by how often query terms appear, weighted by rarity.
BM25 remains the baseline for keyword search because it handles exact matches (SKUs, part numbers, named products) that vector search can miss. Most modern e-commerce search stacks combine BM25 with vector search; see Hybrid search.
- Webhook sync #
-
A data-sync pattern where the store pushes events (product saves, translations, order status changes) to the assistant over HTTPS, rather than the assistant polling or scraping.
Emporiqa uses webhooks exclusively. There is no scraping and no scheduled polling. When you save a product in WooCommerce, Magento, PrestaShop, Drupal, Shopware, or Sylius, the plugin emits a signed webhook event. The assistant receives, validates, and indexes it. Delivery is deferred to the end of the request (or queued async on platforms like Magento and Shopware), so admin saves complete locally while the catalog sync runs asynchronously.
See also: Webhook API docs
- HPOS #
-
High-Performance Order Storage. WooCommerce's modern order-storage backend, introduced as opt-in in WooCommerce 7.1 and default in 8.2+, replacing the legacy custom-post-type storage model.
HPOS stores orders in dedicated database tables, which dramatically improves order-list performance on high-volume stores. Emporiqa's WooCommerce plugin is HPOS-compatible: it reads orders through the abstraction layer and does not assume the legacy post-meta storage model.
See also: WooCommerce integration
- MSI #
-
Multi-Source Inventory. Magento's system for managing stock across multiple physical and virtual warehouses.
MSI lets a Magento merchant have one product available from multiple stock sources (warehouse A, retail store B, third-party fulfillment C). Salable quantity is resolved per source selection. Emporiqa's Magento module is MSI-aware: when a shopper asks about availability, the assistant reports the resolved salable quantity, not the raw per-source stock.
See also: Magento integration
- PrestaShop combinations #
-
PrestaShop's variant system. A combination is a specific intersection of product attributes (e.g. size 42 + blue + cotton).
PrestaShop combinations can have their own SKU, price (specific prices), stock, and image. Emporiqa syncs every combination for a product, so a shopper asking for 'size 42 in blue' gets the exact combination with its specific price and stock, not the parent product with a generic price range.
See also: PrestaShop integration
- Configurable product #
-
Magento's variant-based product type. A configurable has children (simple products), one per attribute combination.
When a shopper on Magento selects size and color, they are adding a specific child simple product to the cart. Emporiqa syncs both the configurable parent and all children, preserving the attribute-to-child mapping so the assistant can answer variant-specific questions (stock, price, availability) correctly.
See also: Magento integration
- Conversion tracking #
-
Attributing store purchases back to chat conversations that contributed to them.
Emporiqa tracks a funnel: chat session started → product added to cart → checkout reached → order completed. When the store emits an order.completed webhook, Emporiqa matches the order to any recent chat session from the same shopper and surfaces chat-attributed revenue in the dashboard. This answers the 'does chat drive sales' question with real numbers.
See also: Features
- CSAT #
-
Customer Satisfaction rating. A post-chat thumbs-up / thumbs-down prompt that captures how the conversation felt to the shopper.
CSAT is the simplest quality signal a chat assistant can collect. Emporiqa prompts the shopper after a conversation (triggered by the agent's `[SHOW_RATING]` tag), records the rating, and flags low-rated conversations for review. Aggregate CSAT appears on the dashboard and lets teams catch regressions before they affect many shoppers.
- Human handoff #
-
Escalating a chat conversation from the assistant to a human agent when the assistant cannot confidently answer.
Emporiqa flags low-confidence answers and sends a handoff request to the store's team. The human agent sees the full conversation history (shopper question, what the assistant tried, products it surfaced, pages it referenced) and picks up without the shopper repeating themselves. Handoff requests include the shopper's email when available so follow-up can happen async.
- Cross-language search #
-
Returning products from the store's catalog regardless of which language the shopper queries in.
A German shopper typing 'warme Jacke für Wandern' can find the insulated hiking jackets you catalogued only in English. The embedding model maps queries and product text into a shared multilingual vector space, so language mismatch does not prevent a match. Requires that the platform plugin supplies multilingual catalog data (WPML or Polylang on WooCommerce, native per-shop content on PrestaShop, store views on Magento, i18n on Drupal, sales-channel language mapping on Shopware, multi-locale on Sylius).
- BYOK #
-
Bring Your Own Key. A pricing model where the customer supplies their own API key for the underlying LLM (OpenAI, Anthropic) and pays the model provider directly.
BYOK shifts LLM costs from the SaaS vendor to the customer, which is attractive at high conversation volumes. Emporiqa does not currently use a BYOK model: LLM costs are included in the per-conversation rate. This keeps billing predictable for stores that do not want to manage API keys.
- Proactive trigger #
-
A rule that automatically opens the chat widget based on shopper behaviour, without the shopper clicking the widget bubble.
Emporiqa supports four trigger condition types: time on page, URL pattern (e.g. open the chat when the shopper reaches /cart or /checkout), pages viewed in the session, and exit intent. Triggers are rate-limited per-session so a shopper does not get bombarded, and each trigger template is configurable in the dashboard, including how long the popup stays visible before auto-closing (5-300 seconds, default 15).
Try It On Your Store
Install free with $25 of signup credit. Connect your products and see how the salesperson handles real customer questions. No card needed at signup. No trial deadline.
$25 credit on signup · No card needed · No trial deadline · Cancel anytime