Recommendations API

Show shoppers what to buy next. Based on what actually sells together.

The Prodx Recommendations API returns grouped product suggestions built from real purchase patterns, product relationships, and shopper behavior. Results come organized by relationship type so the experience is both personalized and merchant-friendly.

Why Generic Recommendations Underperform

Most recommendation engines treat grocery like electronics. They miss context, basket logic, and the relationship between items that are actually bought together.

Weak Cross-Sell Logic

'Customers also bought' often produces noise instead of useful next-best products.

No Relationship Context

Retailers need complements, affinities, impulses, and purchases—not one undifferentiated list.

Limited Merchandising Control

If teams cannot weight recommendation types by page or goal, recommendations stay generic.

How Prodx Structures
What to Recommend Next

Prodx combines the output of multiple underlying engines—Predictions, Complements, Impulses, and Affinities—to deliver grouped, high-impact suggestions that can be tuned by placement and intent.
  • Returns grouped recommendation sets such as Complements, Affinities, and category-specific picks.
  • Uses shopper preferences, category context, purchase history, and product attributes together.
  • Lets retailers tune the recommendation mix to match business goals or page placement.
  • Supports product page carousels, cart suggestions, substitution flows, and discovery surfaces.
  • Surfaces recommendations from grounded data instead of black-box collaborative filtering.
RecommendationsAPI_ image two.

Measurable Impact

When your data infrastructure is clean, every metric improves.

+ %

Increase in daily add-to-cart revenue
Old-site proof point: $2,029/day before vs. $3,310/day after.

+ %

Increase in add-to-cart conversion
Performance improved in an A/B test against a leading AI recommendations engine.

Higher

Basket expansion
Gains were driven by better product selection, not just more products shown.
How a Top-10 Grocer Increased Search Conversion by 35%
See how a leading grocery retailer significantly increased conversion with Clean Data Infrastructure.

Technical Specifications

Every Prodx API is designed for real-world retail environments — high volume, low latency, and continuous data updates. From authentication to uptime guarantees, everything is engineered to integrate cleanly and scale with your business.
Response Format
JSON
Authentication
API key (provided at onboarding)
Rate Limits
Configured per retailer at contract level
Hosting
Prodx-managed cloud infrastructure
Latency
Sub-200ms typical response times
Availability
99.9% + uptime SLA
Data Refresh
Continuous. Product data updates propagate within minutes of catalog changes.
Integration
RESTful API. Standard HTTPS. No SDKs required.
Onboarding
Three weeks typical. Send us your product feed, we handle the rest.
Support
Dedicated account team. Direct engineering access for integration questions.
GET /v1/recommendations?catalogId=GEN_CATALOG &productId=960038380
{
  "recommendationsGroups": [
    {
      "id": "COMPLEMENTS",
      "name": "Complements",
      "recommendationsProducts": [
        { "productId": "960082694", "score": 95 },
        { "productId": "960038416", "score": 88 }
      ]
    },
    {
      "id": "AFFINITIES_ADVANCED_RETAILER",
      "name": "Affinities Advanced Retailer",
      "recommendationsProducts": [
        { "productId": "960050655", "score": 92 },
        { "productId": "196050883", "score": 87 }
      ]
    }
  ],
  "recommendationsRequest": {
    "catalogId": "GEN_CATALOG",
    "productId": "960038380"
  }
}

Ready to see what the Recommendations API can do with your catalog?

Send us your product feed and we'll show you results in weeks.
The data infrastructure everything runs on.
Product data infrastructure for America's leading grocery retailers.
hello@prodx.com
© 2026 Prodx. All rights reserved.