---
title: "What are Aggregations?"
slug: "what-are-aggregations"
updated: 2025-11-05T03:43:34Z
published: 2025-11-05T03:43:34Z
---

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

# What are Aggregations?

# **What is an Aggregation?**

In the context of HG Insights, an **aggregation** is a behavioral data point about a Lead, Contact or Account. It is used **in a behavioral model** and can be **exported directly to your CRM** to help you segment and prioritize your Leads, Contacts and Accounts.

An aggregation can be built to count the******number of active days**, **active users** or **number of specific activities** performed within a given timeframe.

Examples of aggregations you can create in the Data Science Studio:

- **Email_active_days_last_7_days**: counts how many times a person was considered active in the last 7 days
- **Email_shared_project_last_30_days**: counts how many times a person shared a project in your product in the last 30 days
- **Account_active_users_last_10_days**: counts how many users within the account have shown activity in the last 10 days

# **What is an Aggregation used for?**

- As a behavioral data point for a [Likelihood to Buy model.](/v1/docs/lead-engagement-scoring)
- To surface information to your Sales team in the HG Insights likelihood to buy signals field.
- To segment your database when the aggregation is directly pushed to your CRM

![mceclip0.png](https://support.madkudu.com/hc/article_attachments/4410708522637/mceclip0.png)

## Examples of use cases

- The likelihood to buy signals are too noisy for your Sales team because of how often some events are performed by the user.
  - **Solution**: create an aggregation like **email_opened_email_last_90_days** instead in the HG Insights likelihood to buy signals.
- You are launching a new product and would like to identify the people performing specific activities within your current product to surface the people likely to be interested in your new product so that you can start marketing them.
  - **Solution**:
    - create an aggregation like **email_qualifying_activities_last_15_days** which counts the number of these specific activities the person has performed
    - push this value to your Salesforce using HG Insights
    - then send a free trial offer campaign on the leads with a minimum threshold.
- You would like to identify the accounts likely to be interested in your Team plan because of all the different individuals within this company using your free version of the product.
  - **Solution**:
    - create an aggregation like **account_login_users_last_30_days** which counts the number of people within the account who logged in in the last 30 days
    - push this value to your Salesforce using HG Insights
    - build a list of accounts where the aggregation is let's say greater than 5 and send those to your Sales for outreach
- You would like to reward the people who attended at least 3 of your webinars this quarter.
  - **Solution**:
    - create an aggregation like **email_attended_webinar_last_90_days** which counts the number of times the person attended one of your webinars this quarter
    - push this value to your Salesforce using HG Insights
    - then set a personalized marketing campaign on the leads with a minimum of 3 attendance

# **What data is an Aggregation built on?**

Aggregations are built from events coming from your [**behavioral data sources**](/v1/docs/what-type-of-events-can-be-used-in-a-behavioral-model)(product usage, marketing activity, web visits, etc.).

To create aggregations based on behavioral data you collect or on intent data you purchase, HG Insights needs to be able to pull this data from your integrations like Segment, Amplitude, Eloqua, and Marketo (see list of [integrations).](/v1/docs/integrations-overview)

# **How can I create an Aggregation?**

=> Create your own aggregation following [these instructions](/v1/docs/how-to-create-an-aggregation-in-the-data-studio)

-