Splits

Audience: Marketers and analysts who want to compare the effectiveness of different campaign strategies using real audience data.

Prerequisite: Created audience splits in Customer Studio

Overview

Splits measurement charts help you evaluate the outcomes of your marketing experiments across randomized audience groups. This feature provides clarity on campaign effectiveness by comparing performance between control and treatment groups.

Setup steps

1. Split your audience

Before measuring split results, ensure you have completed the following steps in Customer Studio:

1. Created Audience splits.

2. Synced your audience at least once after splits were created.

3. Generated event data (e.g., purchases, clicks) for analysis.

2. Measure results

The Splits chart helps you compare outcomes between audience groups to evaluate the impact of your campaign.

1. Configure chart

In the Charts > Splits view:

• Under Measuring, select the audience you want to evaluate. A tag shows how many split groups exist.
• Use the Measurement window dropdown to define when Hightouch should look for events.
  • Example: “Entry → 30 days after entry”
• Choose a Metric (e.g., Conversions, Revenue). Learn how to create a new metric.
• Optional: Use the Filter by dropdown to refine by user properties or events.

The chart will display performance over time, along with summary statistics.

2. Interpret results

• Lift summary bar (top-right):
  • Green = Significant positive impact
  • Red = Significant negative impact
  • Gray = Not statistically significant
• Solid line: Observed average performance over time.
• Shaded region: 95% confidence interval for each group.

If the confidence bar overlaps 0%, the result is inconclusive.

3. Normalize results

Toggle between:

• Per member (default): Shows average performance per user.
• To holdout group size: Scales results for an even comparison.

You can toggle normalization in the top-right corner. Hovering over the lift card shows raw totals.