How to create a custom audience segment in Streaming Analytics

This feature requires an Enterprise plan on Vimeo OTT. 

You can create custom audience segments to group viewers by the behaviors, attributes, or actions that matter to your business, in addition to the pre-built segments. This article discusses how to build a segment, what each condition type does, and how to choose a time range.

For an overview of how segments work and which pre-built segments are available, see Understanding audience segments in Streaming Analytics

In this article:

• What to know before you create a custom segment
• How to create a custom audience segment
• What do the condition types do?
• Frequently Asked Questions
• Making the most out of custom segments

What to know before you create a custom segment

Before you create a segment, take note of the following:

• Editing: You can’t change a segment’s conditions after you save it. You can only change its name and delivery method. To change the conditions, clone the segment and edit the copy.
• Processing time: Segments refresh once a day at 00:00 UTC. A new segment won’t show any viewers until after the next refresh.
• Signed-in viewers only: Segments only include viewers who were signed in while watching.

How to create a custom audience segment

You can create a segment from scratch or start from a template. To do so:

1. From the Streaming Analytics dashboard, go to Tools > Segments in the left sidebar.
2. Choose how to start:
   1. To build a segment from scratch, click Create in the top-right corner.
   2. To start from a template, click the arrow next to Create and select a template. The form opens with that template’s conditions filled in, and you can change them.
3. Enter a Name and Description for the segment.
4. Add one or more conditions. If you add more than one, choose how to combine them:
   1. AND: Viewers must meet every condition. Each AND condition makes the segment smaller.
   2. OR: Viewers must meet at least one condition.
5. Set a time range for each condition. See Which time range should I use? below.
6. Click Save.

Your segment populates after the next refresh at 00:00 UTC.

What do the condition types do?

Each condition is a rule a viewer must meet to be included in the segment. There are three condition types.

Did have property

Includes viewers based on an attribute or a metric value. You set:

• Property: The attribute or metric, such as count of plays or subscription status.
• Operator: How to compare, such as equals, greater than, or contains.
• Value: The value to compare against.
• Filters (optional): Narrow the condition by device, location, or similar properties.
• Time range: The period the property is measured over.

To compare a viewer against a reference point instead of a fixed value, select Add Comparison and choose what to compare against:

• The same User: The viewer’s own data from another period. Use this to find viewers whose behavior changed.
• All the User base: Your whole audience, such as the average or a percentile across all viewers.
• The segment: Viewers in a specific segment.

Comparisons adjust as your audience changes. A fixed value like “more than 20 hours a month” may describe a smaller or larger share of viewers over time.

Are in percentile

Includes viewers based on how they rank against your audience on a metric. You set:

• Metric: What to rank viewers on.
• Operator: Greater or equal, or less than.
• Percentile: For example, 90 includes the top 10% of viewers.
• Comparison basis: All the User base, or a specific segment.
• Time range: The period the metric is measured over.

Use percentiles for “top” and “bottom” questions. “Top 10% by watch time” always includes about a tenth of your audience, even as overall viewing changes.

Are part of a segment

Includes viewers who already belong to another segment. When the other segment refreshes, your segment updates with it.

You can use this condition to:

• Narrow an existing segment, such as Heavy Streamers in a specific region.
• Combine segments, such as viewers in TV-First Viewers OR Tablet-First Viewers.
• Exclude a group, by combining it with other conditions.

Frequently Asked Questions

Which time range should I use?

Every condition has a time range. The option you choose decides whether the segment updates over time (dynamic) or stays fixed (static)

• Dynamic segments update at every daily refresh, adding and removing viewers as their behavior changes. Use them for groups you want to monitor over time. These options include:
  • Last / Previous 
  • Since 
  • Anytime
• Static segments don’t change after they’re created. Use them for point-in-time analysis, such as a launch week or campaign period, when you want to follow the same group of viewers. When you choose Between and select a specific date range, the segment will not change over time. 

What can I do with a segment after I save it?

After a segment refreshes, it contains a list of the viewers who meet its conditions. You can:

• Apply it as a filter on any dashboard in any of your Hubs
• Download the matching list of User IDs.
• Use it in another segment with the Are part of a segment condition.

To edit a segment’s name, clone it, download it, or delete it, click the three-dot menu (⋮) next to the segment.

Why is my segment failing, empty or smaller than expected?

These are the most common reasons:

• The segment hasn’t refreshed yet: New segments populate after the next refresh at 00:00 UTC.
• Too many AND conditions: Remove conditions one at a time to find the one that excludes the most viewers.
• The time range is too short: Use a longer range so viewers have time to take the qualifying action.
• The conditions conflict: For example, a percentile condition combined with a fixed value that no one in that percentile meets.
• Viewers watch signed out: Segments only include signed-in viewers. Check the Logged Out Viewers metric to see how much of your viewership this affects.

If your segment’s size changes every day, that’s expected for a dynamic segment because its time range moves forward at each refresh. To keep membership fixed, clone the segment and use a Between time range.

Making the most out of custom segments

• Start with one condition and check the segment’s size before adding more. This makes it easier to find which condition causes an empty segment.
• Name segments after the audience, not the logic. For example, “Partial TV Viewers” instead of “TV viewers, 60%+ threshold.”
• Use the description to record why you chose each threshold, since you can’t edit the conditions later.
• Keep segments private while you test them, and make them public once they’re ready for your team.

The examples below show how to combine conditions: