SleekAI reads your live WordPress and WooCommerce data directly. The fiber composition, weave, gsm, and care symbols are usually stored as product attributes or postmeta keys like _fabric_composition and _care_symbols. The data source wizard maps those fields once, then the bot reads them at query time, so the moment you update a product's care label the chatbot answer updates with it. There is no training step and no manual sync.

Yes. If you store care symbol references in postmeta with the symbol image URLs, the bot can include the matching symbol images inline in the reply. Many stores also link to a single care guide PDF per product family, and the bot will deep link to that PDF when the question is broad enough to warrant the long-form answer. The shopper gets the short reply and the option to dive deeper without leaving the page.

Yes. The pattern is identical: store the material composition and care instructions as product attributes, and the bot reads them. Leather bags, wool rugs, cashmere throws, linen tablecloths, and silk pillowcases all benefit because the care nuance is what separates a happy owner from a return ticket. You can scope a single bot to all product types or run one per category with multibot.

The bot is configured to ground every answer in your live catalog, so if a shopper asks about a fabric you do not sell, the bot will say it can only speak to the products in your store. The system prompt explicitly discourages inventing care advice. This keeps the bot honest and keeps your store from being held responsible for advice on someone else's product.

Yes. Display conditions in SleekAI can scope a bot to the order confirmation page (is_order_received_page()) and the My Account orders view, so the first-wash question gets answered exactly when it tends to come up. You can also send a link to a dedicated care page in the post-purchase email and have the bot waiting there with conversation context for that order.

Typical care conversations run two to five turns at roughly 400 to 800 tokens total. With GPT-4o-mini or Claude Haiku 3.5 that comes out to a fraction of a cent per conversation, with Sonnet or GPT-4o it is around two to four cents. Both are dwarfed by the cost of a single avoidable return on a ninety-dollar sweater, which is why care chatbots have a strong unit economics story.

Only if your data says so. The bot reads the care symbols and instructions you have stored against the product. If a product is hand-wash-only, the bot will say so. If you sell items that genuinely require dry cleaning, the bot will quote that. The accuracy comes from the fact that the bot is reading your data instead of generating advice from a general training set, which is where chatbots tend to get this wrong.

Yes. If you sell wool wash, leather conditioner, or stain remover as related products, the bot can mention them when relevant. The system prompt should be configured to do this only when the shopper's question makes the suggestion useful, not as a constant upsell, otherwise the bot loses trust. Conversations that lead to a care-kit add-to-cart are visible in the SleekAI conversation log.