SleekAI reads your product taxonomy (categories, tags, brand), the attributes (color, size, material), and any postmeta you mark as similarity signals. The bot matches on the combination, so a sold-out navy down parka surfaces the next-closest down parka in the same color family and similar warmth rating, not just any winter coat. You can also pin manual substitutions on the product via postmeta and the bot will prefer those.

Only if the price difference is reasonable. The system prompt can be configured with a maximum price differential, often 15 to 25 percent above the original. Above that the bot stops suggesting and instead offers a restock signup, because pushing a 50 percent more expensive product to a price-sensitive shopper damages trust. Conversation logs let you tune the threshold over time.

The bot gracefully falls back to capturing a restock signup for the original SKU if a restock is planned, or simply acknowledging the gap and offering to email when something similar arrives. The system prompt explicitly discourages inventing products. Pretending you sell a similar item that does not exist generates support tickets when the shopper tries to find it, which is worse than losing the visit.

Yes. Display conditions in SleekAI accept any combination of post type, stock status, category, and URL pattern. You can scope this bot to fire only when the product page has zero stock, leaving in-stock pages to a more general support or pre-sales bot via multibot. Each bot has its own system prompt and quick replies.

If only one variation is sold out, the bot can suggest the in-stock variations of the same product first ('navy small and large are in stock, only medium is sold out'). If all variations of the product are sold out, the bot pivots to alternative products. The logic comes from reading the variation-level stock fields in the WooCommerce database, not the parent-level stock state.

Not usually. The chatbot suggests alternatives in conversation, scoped to the OOS context, while your existing related-products carousel keeps running for non-OOS pages. If you want to read your recommendation engine's picks as a data source, SleekAI can be pointed at the recommendation API, but for most stores the catalog and attributes alone produce better OOS-specific suggestions because they account for the shopper's stated intent in chat.

Yes. SleekAI can route a conversation to your support inbox or to a hand-off action that captures the conversation context and the original SKU. For high-value items where a manual ETA check or a customer-service rescue is worth the time, this matters. The conversation context is logged so the human picks up exactly where the bot left off.

The two numbers that matter are alternative-add-to-cart rate (did the suggestion lead to a purchase) and restock signup rate (did the visit at least capture a downstream signal). Both are visible from the SleekAI conversation log joined to WooCommerce order data. Most stores see double-digit recovery on traffic that would otherwise have bounced, which is more than enough to justify the one-time license for any catalog above a few hundred SKUs.