Dimitris Ganotis
dganotis.dev

Commerce Intelligence Engine

A WooCommerce plugin that mines co-purchase patterns from your order history and serves explainable product recommendations on the storefront. Heavy computation runs offline; page requests read from precomputed tables.

GitHub repository•Project site

Overview

WooCommerce's default recommendations are mostly category-based, cross-sells you configure manually, or generic bestsellers. Commerce Intelligence Engine uses your own completed order data to discover what people actually buy together. It builds pairwise associations offline, scores them with support, confidence, lift, and other tunable metrics, and serves them on product pages (and optionally cart/checkout) with optional reason text so the recommendation story stays transparent.

What It Does

  • Mines pairwise product associations from WooCommerce order analytics lookup tables
  • Scores recommendations using support, confidence, lift, margin, stock, and recency
  • Runs full and incremental rebuilds in the background (scheduled or manual)
  • Displays recommendations on product pages, and optionally on cart and checkout
  • Shortcode for manual placement and REST endpoint for headless or custom frontends
  • WP-CLI commands for rebuilds, status, cache flush, and association inspection
  • Admin diagnostics, rebuild logs, and per-product intelligence metabox with pin/exclude overrides
  • Fallback chain: cross-sells → category bestsellers → global bestsellers when mined data is sparse

Key Features

  • Configurable thresholds: minimum co-occurrence, support, confidence, lift
  • Weighted scoring model and category exclusion support
  • Explainable output with reason text (e.g. “Frequently bought together”, “Strong co-purchase signal”)
  • HPOS-compatible; uses WooCommerce order stats and product lookup tables

Installation

In WordPress admin: Plugins → Add New → Upload Plugin. Upload the plugin ZIP, install and activate, then open WooCommerce → Commerce Intelligence to review defaults and run an initial rebuild.

Shortcode

For manual placement:

[cie_recommendations product_id="123" limit="4" context="product"]

product_id is optional (defaults to current product). context can be product, cart, or checkout.

REST API

When enabled in settings:

/wp-json/cie/v1/recommendations/<product_id>?limit=4&context=product

Response includes product_id, recommendations (with score, confidence, lift, reason), source, and generated_at.

WP-CLI

wp cie rebuild
wp cie rebuild --incremental
wp cie status
wp cie flush_cache
wp cie associations <product_id> --limit=10

Requirements

  • WordPress and WooCommerce
  • WooCommerce order analytics lookup tables populated (e.g. wc_order_stats, wc_order_product_lookup). If analytics is disabled or empty, rebuilds complete with zero mined associations and the plugin relies on fallbacks.

Full Documentation

This page is a quick overview. For installation steps, configuration options, rebuild modes, and limitations, see the README in the repository.

Want to support the project?

If this project is useful to you, a coffee helps fund continued work on it.

QR code to Buy Me a Coffee page

Buy Me a Coffee

Support the project by buying me a coffee. Scan the QR code with your phone or click on the link below to buy me a coffee.

Support the project