Sorting

What is Sorting?

Sorting orders a result set by a chosen field or rule such as price, rating, popularity, or newest. Unlike ranking (which blends many signals), sorting applies a clear, user-selected order with predictable tie-breakers.

How It Works (quick)

• Typed fields: Price/date/rating stored as numeric/date doc values for fast sort.
• Stable order: Add tie-breakers (e.g., score → rating → reviews → ID) to avoid jitter.
• UI controls: Dropdown or tabs (Relevance, Price ↑/↓, Newest, Rating).
• URL/state: Expose sort=price_asc in the query string for shareable links.
• Guardrails: Apply hard filters (in-stock/ACL) before sorting; cap page size.

Why It Matters in E-commerce

• Decision speed: Users can zoom in on what matters (budget, quality, freshness).
• Transparency: Predictable order reduces frustration and boosts trust.

Best Practices

• Default to Relevance; keep sort options short and meaningful.
• Normalize currency and units per locale before sorting.
• Use review count as a secondary key for rating sorts.
• Hide price sort for mixed bundles where price is unclear.
• Cache popular sorts; precompute aggregations.

Challenges

• Mixed currencies/units, sparse ratings, popularity bias, and large tie groups.

Examples

• Category page: Relevance default; tabs for Price Low–High, Price High–Low, Rating, Newest.
• B2B: sort by lead time or spec (e.g., torque, wattage).

Summary

Sorting gives users control over order. Keep it fast, stable, localized, and secondary to a strong relevance default.