FAQ

Only products currently online and not snoozed at that location. The latest published menu is checked and disabled products are filtered out.

Yes, via Upsell Overrides — specific products can be pinned per brand and location, and will always be recommended while available, ahead of automatic suggestions. Control is at the individual product level, not category.

No. Recommendations are product-level only; modifiers are never returned.

No. Every suggestion is checked against the live menu and snoozed-product list before it's returned.

Yes — this is the core of how the engine works. It combines AI-based understanding of which products complement each other with evidence from the brand's own order history about what customers actually buy together. A burger leads to a side or drink, not another burger, and which side or drink depends on that brand's own buying data.

No. Recommendations are based on the brand's aggregated order history, not an individual customer's history, and not pooled across brands.

Order history is aggregated at brand level, so all locations share the same behavioural data for ranking. Recommendations aren't segmented by location, city, region or DMA — but availability always is, since a product must be online and unsnoozed at the ordering location.

Unlikely. The engine favours products that complement the basket rather than repeating what's already in it.

It's eligible as soon as its menu is published — availability isn't the constraint. Ranking depends on order history, so a brand-new product won't rank well right away. Use an Upsell Override to promote it immediately.

If there isn't enough history to rank a candidate set, representative menu products are shown instead, so the response is never empty.