Skip to main content

Discount Engine Overview

Architecture​

The discount engine is a deterministic system that applies discounts to cart items based on eligibility rules, priority, and stacking configurations.

Key Features​

  • Deterministic: Same input always produces same output
  • Priority-Based: Lower priority number = higher priority
  • Stacking Support: Multiple discounts can stack if configured
  • Conflict Resolution: Handles mutually exclusive discounts
  • Snapshot Versioning: Immutable discount snapshots for orders
  • Drift Detection: Monitors discount rule changes

Discount Types​

Product-Level Discounts​

Apply to specific products, categories, collections, or tags:

  • PERCENTAGE: Percentage off (e.g., 20% off)
  • FIXED_AMOUNT: Fixed amount off (e.g., ₹100 off)
  • FIXED_PRICE: Set fixed price (e.g., ₹999)

Cart-Level Discounts​

Apply to entire cart subtotal:

  • PERCENTAGE: Percentage off subtotal
  • FIXED_AMOUNT: Fixed amount off subtotal
  • FIXED_PRICE: Set fixed cart total

Tiered Discounts​

Apply based on quantity thresholds:

  • TIERED: Different discounts for different quantities
  • Example: Buy 2 get 10% off, Buy 5 get 20% off

BOGO Discounts​

Buy X Get Y discounts:

  • BUY_X_GET_Y: Buy X items, get Y items free/discounted
  • Example: Buy 2 Get 1 Free

Discount Flow​

Eligibility Rules​

Discounts can have multiple eligibility conditions:

  • Minimum Order Value: Cart must meet minimum amount
  • Product Requirements: Must include specific products
  • Category Requirements: Must include products from categories
  • Collection Requirements: Must include products from collections
  • Tag Requirements: Must include products with tags
  • Customer Group: Only for specific customer groups
  • Usage Limits: Per-customer or total usage limits
  • Date Range: Start and end dates

Priority System​

Discounts are evaluated by priority:

  1. Lower Priority Number = Higher Priority
  2. Priority 1 is evaluated before Priority 10
  3. Highest priority discount wins in conflicts

Stacking Rules​

Stackable Discounts​

When canStack: true:

  • Multiple discounts can apply simultaneously
  • All eligible stackable discounts are applied

Non-Stackable Discounts​

When canStack: false:

  • Only highest priority discount applies
  • Other discounts are ignored

Conflict Resolution​

Mutually Exclusive Discounts​

Discounts can exclude other discounts:

{
  "code": "SAVE20",
  "excludedDiscountIds": ["SAVE30", "FLASH50"]
}

If SAVE20 is applied, SAVE30 and FLASH50 cannot be applied.

Resolution Algorithm​

  1. Sort discounts by priority (ascending)
  2. Build exclusion graph
  3. Resolve mutually exclusive groups
  4. Apply stacking rules
  5. Separate product and cart discounts

Discount Application Order​

  1. Product-Level Discounts: Applied first to individual items
  2. Tiered/BOGO Discounts: Applied after product discounts
  3. Cart-Level Discounts: Applied last to subtotal

Snapshot System​

Discounts are snapshotted when orders are created:

  • Immutable record of applied discounts
  • Used for order history and reconciliation
  • Prevents price changes affecting past orders

Drift Detection​

Monitors discount rule changes:

  • Detects when discount rules change
  • Compares current rules to snapshot rules
  • Generates drift reports for admins

API Endpoints​

  • GET /discounts/calculate - Calculate discounts for cart
  • GET /admin/discounts - List all discounts
  • POST /admin/discounts - Create discount
  • GET /admin/discounts/:id - Get discount
  • PATCH /admin/discounts/:id - Update discount
  • DELETE /admin/discounts/:id - Delete discount
  • GET /admin/discounts/drift-report - Get drift report

Caching​

Discounts are cached in Redis:

  • Cache Key: discounts:rules:{hash}
  • TTL: 1 hour
  • Invalidation: On discount create/update/delete

Performance​

  • Deterministic Engine: Pure function, easily cacheable
  • Redis Caching: Rules cached for fast lookups
  • Batch Processing: Processes all cart items efficiently

Best Practices​

  1. Clear Priority: Use consistent priority numbering
  2. Test Stacking: Verify stacking behavior works as expected
  3. Monitor Usage: Track discount code usage and effectiveness
  4. Set Limits: Use usage limits to prevent abuse
  5. Document Rules: Document discount rules clearly