> For the complete documentation index, see [llms.txt](https://doc.batch.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://doc.batch.com/guides-and-best-practices/batch-ai/how-to-personalize-your-campaigns-with-product-recommendations.md).

# How to Personalize Your Campaigns with Product Recommendations

## What Is the Product Recommendation Score?

Batch AI Predict's **Product Recommendation** score calculates, for each customer in your database, a ranked list of the N most relevant products to recommend. The output is a list of product IDs (configurable: top 1 to top N), based on each customer's purchase history and behavioral profile.

The model is configurable at several **levels of granularity**: SKU, category, or brand (depending on what level you want to personalize in your templates).

{% hint style="info" %}
**Product Recommendation vs. Product Propensity:** which one do you need?

* **Product Recommendation**: you start from a *customer* and want to find the best products to show them. Use this for personalization: *"What should I put in this customer's template?"*
* **Product Propensity**: you start from a *product or category* and want to find the best customers to target. Use this for segmentation: *"Who should I contact for this campaign?"*
  {% endhint %}

## Why Use It?

Generic recommendation blocks (your top 3 bestsellers, the same for every customer) perform significantly below their potential. Each customer has a purchase history and browsing behavior that makes certain products far more relevant than others.

With the Product Recommendation score, you can:

* **Replace static product blocks** with dynamic, individually calculated recommendations.
* **Increase click and conversion rates** by showing customers products they are genuinely likely to buy.
* **Run a single campaign** rather than duplicating templates per segment or category.

{% hint style="warning" %}
**Note on single-purchase customers**: Recommendations for customers with only one past purchase will be less precise, but still more relevant than a generic bestseller list. You can combine this score with a purchase count filter in your targeting if needed.
{% endhint %}

## How to Use This Score in Batch

### Personalization: Dynamic Product Blocks in Your Templates

The core use case: inject each customer's personalized product recommendations directly into your email, push, or in-app templates.

Batch AI Predict exposes the top N recommendations as separate custom attributes on each customer profile:

* `batch_ai_reco_top1`
* `batch_ai_reco_top2`
* `batch_ai_reco_top3`
* (and so on up to your configured N)

**How to set it up:**

1. In your template composer, add a product recommendation block (email carousel, push dynamic content, or in-app product card).
2. Use the Batch personalization syntax to inject the recommended product attributes:
   * `{{ profile.custom_attributes.batch_ai_reco_top1 }}`
   * `{{ profile.custom_attributes.batch_ai_reco_top2 }}`
3. For each product ID, map it to the corresponding product image, name, and URL from your [catalog](https://doc.batch.com/developer/api/cep/catalogs).
4. Each customer who receives the campaign sees a different product selection, calculated specifically for them.

**Recommended test**: Run an [**A/B Content Personalization Test**](https://doc.batch.com/guides-and-best-practices/batch-ai/how-to-measure-the-performance-of-batch-ai-predict-scores): Variant A uses the Batch AI Predict recommendation block, Variant B uses your current logic (static bestsellers, last viewed items, or wishlist). Compare click rates, conversion rates, and RPM. See the Performance Measurement Guide for details.

***

## Real-World Use Cases

| **Industry**              | **Example use case**                                                                                                                        |
| ------------------------- | ------------------------------------------------------------------------------------------------------------------------------------------- |
| **Retail**                | Personalize the "You might also like" block in post-purchase or weekly newsletter emails; every customer sees a different product selection |
| **Media & Entertainment** | Recommend articles, podcasts, or video content in weekly digest emails based on each subscriber's consumption history                       |
| **Travel**                | Recommend the most relevant destinations or experiences based on each traveler's booking history and browsing behavior                      |
| **Beauty & Cosmetics**    | Personalize the product carousel in reactivation emails: each lapsed customer sees the products most aligned with their preferences         |

***

## Related Documentation

* [How to Measure the Performance of Batch AI Predict Scores](https://doc.batch.com/guides-and-best-practices/batch-ai/how-to-measure-the-performance-of-batch-ai-predict-scores)
* [A/B Testing: Message Overview](https://doc.batch.com/getting-started/features/customer-engagement-platform/message/overview#a-b-test)
* [Orchestration Analytics](https://doc.batch.com/getting-started/features/customer-engagement-platform/analytics/orchestration-analytics)
* [Personalization in Batch](https://doc.batch.com/getting-started/features/customer-engagement-platform/message/personalization)
* [Catalogs](https://doc.batch.com/developer/api/cep/catalogs)


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter:

```
GET https://doc.batch.com/guides-and-best-practices/batch-ai/how-to-personalize-your-campaigns-with-product-recommendations.md?ask=<question>&goal=<endgoal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.