Promote Shopify best-sellers in Algolia every Monday

Every Monday at 7am, promote my Shopify best-sellers in Algolia search by creating query rules that boost them for the terms shoppers actually type.

Trigger: cron, Mondays at 7am.

Step 1. Call Shopify List Orders for the past 30 days (lookback window is a knob) and aggregate the line items to identify the top 20 best-selling products by unit volume. Hold on to each best-seller's product title, tags, and vendor, plus the Algolia objectID you use for it in your product index.

Step 2. Call Algolia Analytics: Top Searches over the same window to see the terms shoppers are actually typing, ranked by volume.

Step 3. For each best-seller, reason about which search terms it should show up for. Use the product title, tags, and vendor as hints, and match them against the top-searches list. Prefer terms with real search volume; skip vague, ambiguous, or off-brand matches so you do not over-pin.

Step 4. Before writing anything, use Algolia Get Rule on the objectID you would use (a stable slug per product/term pair, for example bestseller-pin-<productId>-<slug(term)>) to check whether an equivalent pin rule already exists and still applies to the same product. Skip those. Only queue rules that are genuinely new or need to replace a stale pin.

Step 5. Send the batch to Algolia Save Rules (batch) in one call to create or replace the queued query rules, each one pinning the best-seller's objectID in the top slot for the matched term.

Expose these knobs so the ruleset stays tidy: lookback window in days (default 30), top-N best-sellers cap (default 20), max rules per run (default 40), the Algolia index to write to, and an optional exclude list for products that should never be pinned (archived, out of stock, discontinued). Log a short summary at the end: how many best-sellers were considered, how many pins were already in place, and how many rules were written.