Shopware currently provides basic product recommendations such as manual cross-selling and product groups. However, there is no centralized recommendation engine that dynamically generates personalized product recommendations based on customer behavior, business rules, or contextual data. AI influence is also welcome.

Goal

Introduce a Recommendation Engine that automatically suggests relevant products across the storefront based on customer behavior, predefined business rules, product relationships, and contextual factors.

Description

The Recommendation Engine should analyze customer interactions (e.g., viewed products, purchases, cart contents, browsing history, buying history) together with configurable business rules to generate personalized recommendations in real time.

Recommendations should be reusable throughout the storefront and configurable for different placements, such as Product Detail Pages (PDP), category pages, CMS pages, the shopping cart, checkout, and landing pages.

Example Recommendation Types:
Behavioral Recommendations
Customers also bought
Frequently bought together
Recently viewed products
Trending products
Customers with similar interests also purchased

Context-Aware Recommendations
Weather-based recommendations (e.g., umbrellas on rainy days, sunscreen during sunny weather).
Seasonal recommendations (Summer, Winter, Christmas, Easter, Black Friday, etc.).
Time-based recommendations (morning, evening, weekends).
Region-specific recommendations.
Campaign-specific recommendations.

Functional Requirements:
Configurable recommendation strategies.
Rule builder integration to define recommendation logic.
Personalized recommendations based on customer behavior.
Anonymous visitor support using session-based behavior.
AI-ready architecture for future machine learning integrations.
Performance optimized through caching and indexing.
API support for headless storefronts.
Analytics to measure recommendation performance (CTR, conversions, revenue generated).
CMS Block

The block should allow merchants to:
Select one or multiple recommendation strategies.
Configure the number of displayed products.
Define fallback behavior if no recommendations are available.
Customize the layout (slider, grid, carousel).
Reuse the block on landing pages, shopping experiences, and CMS pages.
PDP Integration

The Product Detail Page should support configurable recommendation sections, for example:

Customers also bought
Frequently bought together
Complete the look
Similar products
Accessories
You may also like

Each section should be individually configurable and enabled or disabled.

It should be possible to manage this globally, by category or by product, rather than manually as is the case with the current cross-selling module.

In SW5 there was a "last viewed" function on the product details page.

Would be nice to have this as standard in SW6.

In SW5 there was the function "Customers also bought" on the article detail page.

Would be nice to have this as standard in SW6.