- Which signals matter most (via top‐level weight sliders)
- How deeply each signal probes your catalog (via per-tab configurations)
- Which real-world interactions drive your rankings (through performance and personalization models)
- How business rules can override or augment AI decisions (with merchandising strategies)

- Relevance Configure Gen AI search (multi-model text + image weighting), semantic understanding, and classic text‐processing enrichments. Decide how your catalog is parsed, how out-of-stock items behave, and how deep the AI should dive into meaning versus exact matches.
- Performance Leverage real user engagement with clicks, add-to-cart, purchases, and custom events to train performance-based ranking models. Choose up to three models, assign relative weights to each event, set time decay windows, and let Experro reward the products that truly resonate.
- Personalization Activate AI-powered affinity scoring using global and category-specific attributes, in-session behavior, and historical or cohort-based profiles. Build up to three personalization models, define affinity fields (brand, category, color, etc.), weight your engagement events, and watch Experro tailor results to each shopper’s preferences.
- Ranking Balance all signals together : Relevance, Personalization, Merchandising, and Performance using four top‐level sliders that always sum to 100. Follow industry best practices (e.g., 50–70% relevance, 20–30% personalization, 10–20% merchandising, 1–10% performance) or customize to your brand’s unique goals.