Skip to content

Product Intelligence Explained

clock Avg. 8 min read
·
calendar Last updated on November 19, 2026

Product Intelligence Explained

Introduction

Product Intelligence is Hello Retail's proprietary AI engine. It is the technology layer that powers Search, Recommendations, Pages, Product Agents, Newsletter Content and Retail Media.

Product Intelligence runs alongside the data your store collects about visitor activity, enriching and filling in data gaps where your store's behavioral signals are limited. For new products with no purchase history, long-tail items with sparse data, or stores that are still building up traffic, Product Intelligence provides the product understanding needed to deliver relevant results from day one.

How It Works

Product Intelligence is trained on over 250 million products and the purchasing patterns observed across them. It learns how products relate to each other: which items are alternatives, which are commonly bought together, and which tend to lead to repeat purchases. It does this by analyzing product information (titles, descriptions, images, attributes) alongside aggregated behavioral patterns.

This product-centric approach means the AI understands your catalog at a deep level, even for products that haven't yet accumulated enough visitor activity in your own store to personalize effectively.

What Product Intelligence Powers

Product Intelligence is not a standalone feature you configure in the dashboard. It is the AI backbone that runs behind the scenes across Hello Retail's products.

A visitor's behavioral signals are limited early in a session. They may have viewed only one or two products before running a search. Product Intelligence expands on these signals by inferring broader interests from the products the visitor has engaged with. For example, if a visitor has looked at a trail running shoe, Product Intelligence uses patterns associated with that product type to rank search results toward running gear, technical fabrics, and outdoor categories.

Product Recommendations

Product Intelligence determines how products relate to each other: which items are commonly bought together, which are alternatives, and which tend to appear later in a customer's purchase journey. When a visitor views a coffee machine, recommendations can surface the specific accessories customers with that product tend to buy next, such as coffee beans, filters, or a grinder, rather than relying on simple co-view signals.

Pages

Product Intelligence influences how products are ranked on category and brand pages for each visitor. A visitor who has shown interest in premium brands will see those products ranked higher within a category. Two visitors on the same category page can receive a different product order based on their interests.

Product Agents

Product Agents use Product Intelligence to select which products to include in automated emails. When a customer has purchased a product, the system uses product clustering to identify complementary items that customers with a similar purchase tend to buy next - and surfaces those in the follow-up email rather than generic bestsellers.

Newsletter Content

Product Intelligence determines which products to include in newsletter recommendations for each recipient. Rather than sending the same product selection to your full list, the system selects products based on each recipient's purchase history and the product relationships it has learned.

Retail Media

When a Retail Media campaign promotes a brand or product attribute, Product Intelligence identifies which products in your catalog best match the campaign criteria for each visitor. Promoted products are inserted based on relevance, not just position availability.

Product Lookup

Product Lookup lets you inspect how Product Intelligence has understood a specific product: its assigned clusters, tags, and relationships to similar products. See Editing Clusters and Tags for details on how to view and adjust these values.

Available Insights (API)

For technical users, the Product Intelligence API exposes analytical data about individual products via a GraphQL endpoint. You submit a product URL and receive insights based on the product and similar products grouped together, giving the system a larger and more reliable dataset to draw from.

The API can also be used to push Product Intelligence data into third-party systems such as CDPs, ESPs, and PIM systems, making the insights available across your wider marketing and commerce stack.

Available data points include:

Insight Description
Product Lifetime Value Expected further revenue from customers who buy this product
Retention Rate Percentage of buyers who return to purchase something else
Recurrence Whether the product is typically repurchased, and the probability of repurchase over time (weekly breakdown)
Recurring Period If there is a strong repurchase pattern, the peak period (in weeks) when customers tend to reorder
Quarterly Sales Insights into when people tend to buy the product throughout the year
Hourly Sales The typical time of day when the product is purchased
Discount Distribution Probability of purchase at different discount percentage ranges, showing how price sensitivity varies
Budget Zones A classification of products into 5 price levels, from Cheap to Expensive

The data is made available in a GraphQL API that is available from the URL: https://core.helloretail.com/pi/graphql.

Authentication requires either an active session on my.helloretail.com or an API key. Full schema documentation and a GraphiQL test interface are available at the endpoint URL when authenticated.

For detailed API usage, query examples, and authentication setup, see the Product Intelligence API documentation.

Requirements

Before insights can be generated for a product:

  • The product must exist in Hello Retail (typically through your product feed).
  • The product must have a description and a valid image so it can be matched to similar products.
  • There is a processing delay after new products are created before mapping completes.

Not all products will return insights. If there is insufficient data to produce accurate results, the API returns an empty response rather than unreliable data.

Q&A

Does Product Intelligence require any setup or configuration?

No. Product Intelligence runs automatically once your store is connected and your product feed is active. It powers Search, Recommendations, Pages, Product Agents, Newsletter Content and Retail Media without additional configuration.

Does Product Intelligence replace my store's behavioral data?

No. Your store's own visitor behavior (clicks, views, purchases) is always the primary signal for personalization. Product Intelligence runs alongside it, enriching results and filling in gaps, especially for new products, long-tail items, or stores still building up traffic. The more data your store accumulates, the more it drives personalization directly.

Can I see how Product Intelligence understands a specific product?

Yes. Use Product Lookup in the Hello Retail dashboard to view and edit the Product Intelligence clusters and tags attributed to your products.

Is the Product Intelligence API included in all plans?

Contact your Hello Retail account manager or support@helloretail.com for API access and plan details.