> For the complete documentation index, see [llms.txt](https://docs.dinmo.io/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.dinmo.io/intelligence/event-propensity.md).

# Event Propensity Scoring

Estimate the likelihood that a profile or subscription will perform a defined business event within a future window. Use the scores to prioritize eligible audiences for conversion, engagement, or retention.

## Choose a scoring use case

| Business need                                | Mode and scoring grain                                                    | Example target                              |
| -------------------------------------------- | ------------------------------------------------------------------------- | ------------------------------------------- |
| Prioritize likely purchasers                 | Generic event propensity; one score per profile                           | `order_completed`                           |
| Identify registration or trial opportunities | Generic event propensity; one score per profile                           | `registration_completed` or `trial_started` |
| Prioritize content engagement                | Generic event propensity; one score per profile                           | `premium_content_viewed`                    |
| Anticipate subscription cancellation         | Subscription lifecycle propensity; one score per profile and subscription | `cancellation_requested`                    |

Generic mode learns from the timing and mix of behavioral events. Subscription mode also reconstructs subscription state and history, with optional behavioral signals such as usage or content consumption.

For retention, choose a target that leaves time to act. A cancellation request can be more actionable than the effective subscription end. Subscription mode requires mappings for starts, cancellation requests, and ends even when only one movement is the prediction target.

## How it works

Training uses historically observable subjects from the configured event streams. The current User model defines who can receive a score; it does not filter that historical training population.

After training, create scoring configurations for the eligible populations you want to serve, with their own refresh schedules. The interface calls these configurations predictions.

Outputs include a score, a training-reference percentile, a score band, up to three signed local reason families, and scoring history. Interpret a score as a probability only when its score semantics confirm calibration. Percentiles and reasons support prioritization and interpretation; they do not prove that a campaign will change the outcome.

If you need to choose which individual product or offer to show, use [Product Recommendations](/intelligence/product-recommendations.md). Event Propensity predicts an event, not an item ranking.

## Documentation guide

| Start here                                                                                           | What you will learn                                                                                        |
| ---------------------------------------------------------------------------------------------------- | ---------------------------------------------------------------------------------------------------------- |
| [Prepare your data](/intelligence/event-propensity/prepare-data.md)                                  | Technical prerequisites, raw event mappings, subscription history, input limits, and readiness thresholds. |
| [Configure and refresh](/intelligence/event-propensity/configure-and-train.md)                       | Target and horizon choices, training, scoring populations, and refresh schedules.                          |
| [Understand performance and scores](/intelligence/event-propensity/understand-scores-and-reasons.md) | Temporal evaluation, probability, percentile, signed reasons, and warehouse outputs.                       |
| [Use results and best practices](/intelligence/event-propensity/limits-and-best-practices.md)        | Actionable targets, leakage risks, audience selection, experiments, and monitoring.                        |

Event Propensity Scoring must be enabled for your workspace. Confirm access with your DinMo administrator or customer success contact.


---

# 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://docs.dinmo.io/intelligence/event-propensity.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.