Yes, if you tag your products with a 'support type' attribute or taxonomy term, the bot reads it directly. Most run-specialty shops already maintain this attribute in WooCommerce because they filter the category page by it. SleekAI uses the same field. The bot can recommend three neutral options under 250g for a runner who said they are neutral with a 10k goal, and it will not surface stability shoes that do not match the runner's profile.

Width is usually a WooCommerce variation along with size, and SleekAI reads both. A runner asking for a wide 11.5 in a specific model gets a yes-or-no answer from real stock rather than 'check our stock page'. For brands that sell standard and wide as separate SKUs rather than variations, the data-source wizard can map those into the same logical product so the bot still answers cleanly.

If your trail-shoe products are tagged with terrain attributes (technical, fire road, mixed) and you store lug depth or upper material in ACF, SleekAI reads those fields and matches them to the terrain the runner describes. A runner mentioning steep loose descents gets a 4mm-or-deeper lug recommendation from your real catalog rather than a generic 'try a trail shoe' deflection.

Carbon-plate and stack-height rules vary by federation and event, so the bot will not give legal advice on whether a specific shoe is permitted in a specific race. It can quote the manufacturer-stated stack height from your product data and recommend the runner verify with the race organisation. That keeps the answer accurate and removes any liability around a federation rule change.

The bot can collect the relevant context (current shoe, where it hurts, weekly mileage, when it started) but should not give medical advice. The configured system prompt steers the runner toward booking a fit session with a staff fitter or seeing a sports physio. The captured context is in the chat log so the in-store fitter sees the runner's history without re-asking the questions on arrival.

With LatePoint, Bookly, Amelia, or a similar plugin, SleekAI can quote real slots and link the runner to the booking page with the gait service preselected. With a small custom REST endpoint, the bot can create the booking directly inside the chat. Most stores start with the linked flow because it requires no dev time, then add direct booking once chat volume justifies it.

Yes. Multibot lets you run a road-shoe-focused bot on road category pages, a trail bot on trail pages, and a race-shoe specialist on the race-day collection. Shared shop info (hours, returns, fit sessions) lives in a common knowledge pool so policy edits happen in one place. Display conditions cover taxonomy, URL pattern, role, and 28+ other rules.

All logs stay in your WordPress database with the model name, token usage, and the page URL the chat happened on. Nothing routes through a Sleek-hosted server, so the runner's mileage, injuries, and current shoe stay on your infrastructure. That matters for shops doing email follow-up or run-club nurture campaigns because the chat context is right there alongside the customer's purchase history in WordPress.