Storzen — AI Recommends
for WooCommerce

What is Storzen AI Recommends?

Storzen AI Recommends for WooCommerce brings intelligent, conversion-boosting product recommendations to your store without any external API calls or subscriptions. The AI engine runs entirely on your server using your own order history — no account, no API key, no third-party dependency.

It is built as a fully modular system: each recommendation surface (Frequently Bought Together, Related Products, Cart, Post-Purchase, Recently Viewed) can be enabled or disabled independently. Social proof badges and low stock urgency badges layer on top of every surface automatically.

System Requirements

WooCommerce is a hard requirement — the plugin checks for it on activation and deactivates itself with an admin notice if it isn't active. Elementor is entirely optional; the widget only registers if Elementor is detected.

Installation

Method A — WordPress Plugin Directory (Recommended)

Method B — Manual Upload

Quick Start

After activation, all core modules are enabled by default. The plugin starts learning from orders immediately — no manual setup required for basic usage.

• The Frequently Bought Together widget appears on single product pages automatically. Go to Settings → FBT to configure position, max products, and bundle discount.
• Related Products AI replaces WooCommerce's default related products on single product pages. Go to Settings → Related Products AI to pick the scoring mode (co-purchase, taxonomy, or hybrid).
• Cart Recommendations appear on the cart page and mini-cart automatically. Go to Settings → Cart to toggle each placement independently.
• The Recently Viewed widget appears on product pages by default. Place it anywhere else using the [sprwoo_recently_viewed] shortcode or Gutenberg block.
• Post-Purchase Recommendations show on the order thank-you page automatically — no shortcode needed.
• Go to Settings → Social Proof and Low Stock to fine-tune badge thresholds so new stores never show misleading data.
• Click Rebuild co-purchase matrix on the Dashboard after activation if your store has existing order history to seed recommendations immediately.

Frequently Bought Together

An Amazon-style "customers also bought" widget displayed on single product pages. It's powered by the co-purchase frequency matrix — the products that most frequently appear alongside the current product in completed orders are ranked to the top. On stores with sparse order history, the taxonomy-based fallback provides relevant companion suggestions immediately.

Bundle discount

When enabled, the FBT widget shows a single Add All to Cart button and a discount line showing the savings. The discount can be configured as a percentage (e.g. 10% off) or a fixed currency amount. Both the button label and the section title are fully customizable.

Display position

Choose where the FBT widget appears relative to the product summary:

Related Products AI

Replaces WooCommerce's default random/taxonomy-based related products section with AI-ranked suggestions tailored to each product. The override uses the official woocommerce_related_products filter — no template files are modified, making it compatible with any correctly coded WooCommerce theme.

Scoring modes

Recommendation scores are cached as WordPress transients with a 1-hour TTL, so the scoring query runs at most once per hour per product — frontend page load time is unaffected after the initial cache warm-up.

Cart Recommendations

Displays contextual product suggestions at two points in the cart experience: the full cart page and the mini-cart sidebar. Products are scored in real time from the current cart contents and items already in the cart are automatically excluded from suggestions.

Placements

• Cart page — rendered below the cart table using the woocommerce_after_cart_table hook
• Mini-cart — rendered inside the mini-cart widget using the woocommerce_before_mini_cart_contents hook

Each placement can be toggled independently under Settings → Cart. The section title and maximum number of recommended products are also configurable.

Recently Viewed Products

Tracks which products a shopper has browsed using the browser's built-in sessionStorage — no cookies, no server-side queries for tracking, and no data stored in your database. The list clears automatically when the browser tab closes.

Automatic placements

• Product pages — the Recently Viewed widget appears below the product content automatically (toggle via Settings → Recently Viewed → Show on product pages)
• Cart page — optionally display recently viewed products on the cart page (toggle via Settings → Recently Viewed → Show on cart page)

Shortcode

[sprwoo_recently_viewed max="4" columns="2" title="Your browsing history"]

Post-Purchase Recommendations

Shows a "Complete your purchase" recommendation widget on the WooCommerce order-received (thank-you) page — catching customers at their highest-intent moment, immediately after they've just bought something.

How product seeding works

The module identifies the highest-value item in the completed order and uses it as the anchor product for the recommendation query. Products that were just purchased are excluded from the results so shoppers only see genuinely new suggestions. The widget renders using the same AI scoring engine as the rest of the plugin (hybrid mode by default).

Display controls

Under Settings → Post-Purchase you can configure the section title, maximum number of recommendations, column count, and whether to show price, star rating, and Add to Cart button on each card.

Social Proof Badges

Social proof badges appear automatically on recommendation cards across all surfaces — FBT, related products, cart, post-purchase, recently viewed, Gutenberg block, and Elementor widget. All data comes from WooCommerce; no third-party lookups or external services are involved.

Badge types

Choose which badge type to display under Settings → Social Proof → Badge type. Minimum thresholds for purchases, reviews, and sales can all be set independently — so stores with sparse data never show badges with misleading low numbers.

Low Stock Urgency Badges

Shows urgency badges on recommendation cards when a product's WooCommerce stock level falls at or below your configured threshold. The badge renders before social proof badges on all recommendation surfaces.

Badge styles

Variable product handling

For variable products, the badge checks all available variations and uses the lowest remaining stock count across them. Products that don't have WooCommerce stock management enabled are never shown a low stock badge, regardless of the threshold setting.

Analytics Dashboard

A built-in analytics dashboard inside WP Admin tracks the performance of every recommendation widget — without sending any data outside your server.

Tracked events

• Impressions — a recommendation widget was rendered and visible on the page
• Clicks — a shopper clicked a recommended product card
• Add-to-cart — a shopper clicked Add to Cart on a recommendation card

Events are stored in the {prefix}sprwoo_analytics database table and can be filtered by 7, 30, or 90-day windows. The table auto-purges rows older than your configured data retention window to keep your database lean.

Gutenberg Block

A native WordPress block with server-side rendering lets you place the Smart Recommendations widget on any page or post without code. Find it in the block inserter by searching for "Smart Recommendations" or "Storzen".

Block controls (in the editor sidebar)

• Recommendation type — choose from related, cart-based, or recently viewed
• Max products — how many recommendation cards to render
• Columns — grid column count (1–4)
• Section title — heading text displayed above the widget
• Show price — toggle product price visibility on cards
• Show rating — toggle star rating visibility
• Show add to cart — toggle the Add to Cart button on cards

Elementor Widget

When Elementor is active, a Smart Recommendations widget appears in the Elementor panel (search keywords: recommendations, products, woocommerce, storzen, AI). It renders the full recommendations widget with a live canvas preview and additional style controls that go beyond what the shortcode offers. The widget only loads at all if Elementor is detected as active — it adds zero file weight to sites that don't use Elementor.

Style controls

• Title color — customize the section heading color independently of the theme
• Title spacing — control the gap between the heading and the product grid

Shortcode

For placing other recommendation types anywhere on your site without code, use the Gutenberg block or the Elementor widget, both of which support any recommendation type.

All Settings

Every option, grouped by the settings tab it lives under (Storzen Reco → Settings in the admin menu). All settings are stored as a single serialized option under the key sprwoo_settings.

Frequently Bought Together

Related Products AI

Cart Recommendations

Recently Viewed

Post-Purchase

Social Proof

Low Stock

Analytics

How the AI Works

The engine builds a co-purchase frequency matrix from your completed WooCommerce orders. Every time two products appear together in a single completed order, their co-occurrence count is incremented in the {prefix}sprwoo_copurchase table. Products bought together frequently accumulate a higher score and rank higher in recommendations.

Matrix updates

The matrix updates incrementally with every new completed order via the woocommerce_order_status_completed hook — there's no batch job to schedule. On first install, click Rebuild co-purchase matrix from the Dashboard to process your historical order archive.

Taxonomy fallback

On new stores with no order history, a taxonomy-based scoring fallback calculates relevance from category and tag overlap between products. This ensures recommendations are meaningful from day one, even before any sales data exists.

Scoring modes summary

Performance

Scoring results per product are cached as WordPress transients with a 1-hour TTL. After the initial warm-up, the AI query never runs on a live page load. The flush_scoring_cache() method is called automatically after a matrix rebuild to force fresh scores.

Privacy & Data

Storzen AI Recommends does not collect, transmit, or share any data with external servers. All scoring, storage, and analytics run entirely within your WordPress database. No account or API key is required at any point.

What the plugin stores

Recently viewed — no server storage

The recently viewed tracking uses the browser's built-in sessionStorage exclusively. No data is written to the server for this feature. The data is temporary, client-side only, and is automatically cleared when the browser tab closes.

Frequently Asked Questions

Changelog