Knowledge Center

Knowledge Center > Insights > Developer Reference > Insights Calculated Field Patterns

Insights Calculated Field Patterns

Overview

This article outlines 11 common patterns for deriving Calculated Fields in Insights. These patterns have proven generally effective in illuminating actionable insights from subscriber data and creating useful customer segmentation. 

Terminology

  • Calculated Fields: Any field that Insights calculates using the patterns outlined in this article is called a Calculated Field. Calculated Fields can be either metrics or attributes. For example, you may feed an event into Insights every time one of your subscribers views a video on a webpage. You can use the Count metric pattern (described below) on those events to create a metric called "Videos Viewed per Day".
  • Primary Fields: Any Account or User data that you send directly into Insights from your other systems that you want to use as an Account or User field in Insights are referred to as Primary Fields. For example, your product may already count the number of videos each subscriber views each day. You can send that number into Insights every day as a daily primary metric called "Videos Viewed per Day". The field patterns described in this article do not apply to Primary Fields.

Metric Patterns

Metrics are fields in Insights that quantify an Account or User's behavior at a repeating time interval -- either daily, weekly, or monthly. Metrics are always numeric. The following metric patterns are common in Insights.

 

Pattern

Description

Examples

Counts

A calculated metric that counts either event, user, or session records stored in Insights.

  • Articles Viewed per Day (Account-level)

  • Articles Viewed per Day (User-level)

  • Monthly Active Users (MAUs)

Averages

A calculated metric that averages a primary metric or primary attribute over a period of time.

  • Average Logins per Day (Last 30 Days)

  • Average Watching Time (Last 30 Days)

  • Average Session Duration (Last 30 Days)

Percentiles

A calculated metric that indicates the percentile rank of an Account or User against all Accounts or Users for a particular primary attribute or metric.

  • Account Age Percentile

  • Article View Count Percentile

Percent Changes Indicates the percentage a metric has change in a recent amount of time.
  • Daily Active Users (% change last 30 days)
  • Videos Viewed per Day (% change last 30 days)

Ratios

Compares the value of one metric with another.

  • Photo Uploads per Session

  • Videos Viewed to MRR Ratio

Sums

Sums primary metrics for Accounts or Users stored in Insights.

  • Total Clicks Last 30 Days

  • Total Calls (MTD)

  • Advertiser Logins (All-time)

Trends

Indicates if a metric is trending up, down, or flat.

  • Video Viewing Trend
  • New User Adoption Trend

Attribute Patterns

Attributes are fields in Insights that tracks a characteristic of an Account or User. Attribute can be numeric or text and differ from metrics in that they do not update on a predetermined frequency; in fact, many attributes update quite infrequently (or not at all). The following metric patterns are common in Insights.

Pattern

Description

Examples

Coalesce

A calculated attribute that takes two or more attributes from different data sources that represent the same value and prioritizes which one to use.

  • Account Name

  • Account Zip Code

  • Account Industry

  • Account Annual Revenue

Highest Value A calculated attribute that indicates the highest value for a particular Account or User field.
  • Highest Payment Amount

Most Frequent Value

A calculated attribute that indicates the most popular event attribute value for an Account or User over a particular time period.

  • Favorite Browser (last 30 days)

  • Favorite TV Series (last 30 days)

Most Recent Value

A calculated attribute that indicates the most recent event attribute value for an Account or User.

  • Most Recent Payment Date

  • Most Recent Payment Amount

  • Last Comment Date

  • Most Recent Event

Last modified

Tags

This page has no custom tags.

Classifications

(not set)