Frequently Bought Together mining. Push profit-ranked bundles in one click.

MarketBasketAnalysis turns your real BigCommerce order history into ranked bundling and cross-sell opportunities. It mines what your customers actually buy together, scores every pairing by support, confidence, lift, and margin contribution, and lets you act on the winners in one click: a native cart-discount bundle promotion, a bundle product, or a storefront widget.

SEVEN MINING ENGINES

Every store gets a full mining stack, not one black-box model: co_occurrence (fast baseline pair counting, runs on every store), sql_pairs (SQL-native pair mining tuned for large order volumes), fp_growth (full frequent-itemset mining for 3-plus item patterns), seasonal_cohort (surfaces pairings that spike in specific seasons), return_aware (discounts pairings that come back as refunds), ai_catalog (LLM-assisted cold-start pairings for new or low-order stores, using your own Anthropic or OpenAI key), and hosted_hui on Plus (high-utility itemset mining on our hosted engine that ranks bundles by profit contribution, not just frequency).

Every recommendation is inspectable. You see the support, confidence, lift, and revenue math behind each suggestion, and you can pause, suppress, or pin any rule.

FROM INSIGHT TO REVENUE IN ONE CLICK

The Opportunities grid ranks co-purchase rules with a Why panel that explains each recommendation in plain numbers. The one-click Bundle action creates an automatic BigCommerce cart-discount promotion, with optional AI-generated bundle copy. Push to product promotes an opportunity into a native BC product. The Weekly plan is a prioritized to-do list of merchandising actions, refreshed from your latest mining run.

STOREFRONT WIDGETS, NO THEME EDITS

All three widgets install through the BigCommerce Scripts API and remove themselves cleanly on uninstall: Frequently Bought Together on product pages, Bundle Builder (a customer-facing multi-item bundle picker), and Cart Drawer recommendations. Catalyst and headless storefronts use the same recommendations through a stable JSON API.

BUILT FOR OPERATORS AND AGENTS

AI merchandiser chat lets you ask questions about your own mining data. A public REST recommendations API plus an MCP server means AI agents and custom integrations can query your store's intelligence. Diagnostics show live activity: every mining run and webhook, with customer data redacted before it is recorded. AI catalog enrichment generates product descriptions and bundle copy with your own provider key and writes them back in one click.

PRIVATE BY DESIGN

Only line-item order data is read. Customer PII never reaches the recommendation engine. AI features are optional and bring-your-own-key: your catalog data goes directly from our backend to your provider using your credentials.

PRICING

Free ($0): the local mining engines (co_occurrence, sql_pairs, fp_growth, seasonal_cohort, return_aware, ai_catalog with your own AI key), the opportunities grid with full rule math, storefront widgets, and one-click bundle promotions.

Plus ($99 per month, or $948 per year, about $79 per month): adds the hosted profit-weighted mining engine (hosted_hui), the public REST recommendations API and MCP surface, the AI merchandiser chat, catalog enrichment, and higher scan caps. Plus is enabled through our team today: contact us at

and we turn it on for your store.

Enterprise and hosted engine: custom pricing, contact us.

There is no charge at install. The Free tier is real and permanent.

SUPPORT

Docs:

. Support:

. Site: