---
title: "Predictive Segments: Discount Affinity"
slug: "predictive-segments-discount-affinity"
description: "Understand Discount Affinity in Insider One Predictive Segments to identify users likely to purchase with discounts based on past behavior."
tags: ["discount affinity", "predictive segmentation", "predictive segments"]
updated: 2026-04-12T18:00:39Z
published: 2026-04-12T18:00:39Z
canonical: "academy.insiderone.com/predictive-segments-discount-affinity"
---

> ## Documentation Index
> Fetch the complete documentation index at: https://academy.insiderone.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Predictive Segments: Discount Affinity

Offering discounts is one of the quickest ways to attract potential customers and encourage purchases. While some customers prefer to maintain a premium shopping experience, others are more responsive to discounted offers. Identifying and targeting users with an affinity for discounted products is valuable for marketers, as it allows them to personalize campaigns that drive conversions without sacrificing margins.

Discount Affinity uses a multi-class [classification model](/v1/docs/insiders-approach-to-managing-ml-tasks) to segment users based on their sensitivity to discounts, classifying them into three groups: **High**, **Medium**, and **Low** affinity. This model helps you tailor promotional offers, such as sending heavy discounts to high-affinity users or avoiding discounts for low-affinity users likely to buy at full price.

![](https://cdn.document360.io/c6df4583-da94-4cb2-bb8a-be0cbdd11109/Images/Documentation/Screen%20Shot%202022-01-18%20at%2015.49.59.png)

The Discount Affinity segment aims to determine a user's likelihood of engaging with discounted products by analyzing their past interactions and purchasing behavior.

## Model Details

The Discount Affinity model runs on a behavior-based classification algorithm. The following details explain how the model operates and how frequently user segments are updated:

- **Model type**: Multi-class classification
- **Segment update frequency**: Daily
- **Data span**: Last 6 months
- **Calculation time**: 30 days

## Data Used

The model evaluates a user's interactions with discounted products using data from both web behavior and CRM sources, where available. All session-based clickstream events that include product discount information are considered, such as:

- Product page views
- Add-to-cart events
- Purchases

> CRM purchase data is included only if combined with web session data, as in-session behavior forms the basis of this segmentation. CRM-only setups are not used for this segment.

To ensure accurate discount calculations, the **unit_price** and **unit_sales_price** parameters must be correctly implemented.

## Key Features

- Past user interactions with discounted items
- Purchase behavior during promotions or seasonal campaigns
- Browsing patterns indicating price sensitivity (e.g., visiting sale sections or filtering by price)

## Segment Breakdown

Users are segmented into the following categories:

- **High Discount Affinity** (Discount Seekers): These users consistently engage with discounted products, browsing sale items, adding them to their carts, and making purchases during sales events. For example, a user who frequently visits the sale section, browses discounted products, and makes purchases during promotional periods.
- **Medium Discount Affinity**: Users in this category show moderate engagement with discounted products. They might occasionally purchase during sales, but don't rely heavily on discounts. For example, a user who typically purchases full-price items but uses discounts for higher-value products.
- **Low Discount Affinity** (Premium Shoppers): These users rarely interact with discounts and are more inclined to buy full-priced or newly released products. For example, a user regularly purchases from the new arrivals section and does not engage during sales campaigns.

> [!NOTE]
> By default, new users with no prior interaction data are assigned to the Medium affinity group. As their behavior develops, they are reclassified accordingly.

## Use Cases

This segmentation enables targeted promotional strategies, such as:

- Offering exclusive discount codes to high-affinity users
- Preserving brand value by limiting discounts for low-affinity users
- Testing discount elasticity across user groups to maximize ROI

## Related

- [Use Predictive Segments](/use-predictive-segments.md)
- [FAQ about Predictive Segments](/faq-about-predictive-segments.md)