On the My Account page, the customer is already logged in, so SleekAI can scope queries to that user's orders only via display conditions. For email-based lookups outside login, the bot asks for order number plus email or zip code as a soft verification step. Sensitive details like full address are not surfaced; the bot quotes status and dates only. For full account changes, the conversation hands off to a human.

Most stores store the linked PO ID and vendor ETA as postmeta on the order line item or the product. SleekAI's data source wizard maps those fields once. Larger stores often have a separate purchase orders custom post type or a connected ERP; the bot can be pointed at the WordPress-side mirror of that data. The bot reads whatever shape your replenishment system writes to WordPress.

By default the bot informs and routes a cancellation request to a human. If you trust the bot to act, you can configure it to call a WooCommerce REST endpoint to cancel an order when the customer confirms. Most stores prefer the routed flow for cancellations to keep a human in the loop on the refund. Either configuration is supported through the chatbot's action layer.

The bot reads each line item separately and quotes the status per line. A typical answer for a three-line order is 'Line 1 shipped on October 5 with tracking X, line 2 is on backorder ETA November 4, line 3 is awaiting your color choice on the customizer'. Each line's status is independent and the bot summarizes the slowest one as the current order ETA, which is the question most customers actually ask.

Once an order has shipped, tracking numbers usually live on the order meta. The bot can quote the tracking number and a current scan status if you store the scan event in WordPress. For deeper integration, the bot can call a courier API on demand and quote the latest scan. Most stores find that quoting the tracking number plus the latest event from a daily polled cache is enough to answer the recurring 'where is it' questions.

Honestly, no, and pretending otherwise damages trust. The bot quotes the current ETA from your data and the date that ETA was last updated. If the data was updated more than a week ago, the bot can note that and offer to escalate. The signal customers need is honesty about the current best estimate, not a probabilistic guess that turns into a chargeback when it is wrong.

Materially yes. Status tickets are the largest single category of tier-one support volume for stores with any meaningful backorder rate. Conversation logs in SleekAI show how many status checks were resolved without a human and which queries had to be escalated. Most stores see double-digit percentage reductions in status tickets within the first month, with the savings compounding as the bot's data mappings get refined.

Two options. The light-touch path is a nightly sync that mirrors PO and ETA data to a WordPress custom post type or postmeta, which the bot reads. The deeper path is a WordPress endpoint that proxies a live call to the ERP when the bot needs PO data, with a short cache to control load. Either works; the choice depends on how often your ERP data changes and how much load the live calls would add.