Catalogs and menus

Create Your First Catalog

Add a clear catalog name, categories, products, prices, images, and tags, then publish a useful first version.

What this helps you do

Build a small catalog that customers can browse and your AI can use.

Create Your First Catalog visual example

Start with one customer job

Choose what this catalog should help a customer do: browse a menu, compare products, review service packages, inspect exhibition items, or find the right product before contacting you.

Use a customer-facing title such as Menu, Product guide, Collection, Services, or Equipment. The public chatbot can use this same title for its catalog button.

Set the basic information

Choose the currency that customers expect, then write one short introduction. The introduction should explain what is in the catalog, not repeat a long company history.

Keep the catalog name under 80 characters and the introduction under 260 characters so the heading remains readable on a phone.

Create categories before adding many items

Use a small number of clear groups such as Main dishes, Drinks, Lighting, Components, Service plans, or New collection. Customers should understand each category without internal company knowledge.

Avoid categories that contain only one item unless that separation is genuinely useful.

Write each item for a real question

Add the product name, price when appropriate, category, description, tags, image, and official detail link. Explain what the item is, who it fits, the most important difference, and any limit customers need to know.

Use tags for searchable attributes such as material, use, size, style, compatibility, or dietary option. Do not copy the same vague marketing sentence into every item.

Save, preview, and publish

Save while you work, then preview the public page on a phone. Check the catalog title, filters, images, prices, detail sheet, main chatbot link, and mini assistant.

Turn on the public catalog only after at least one visible item has a name. Publishing updates the public page and queues AI training only when training-relevant catalog content changed.

Quick checklist

The catalog has one clear customer job?
The title and introduction fit on a phone?
Categories use customer language?
Every visible item has a useful name and image?
Prices and currency are correct?
The public preview and mini assistant were tested?