InsightsAudience Trends

Audience Trends

The Audience Trends dashboard tracks how your audiences change over time. Monitor audience size, growth and shrink rates, membership change velocity, and historical snapshots to understand audience dynamics and measure the impact of campaigns, product changes, and market shifts.

Every time an audience is evaluated (during a sync run or a scheduled evaluation), SignalSmith records the member count and a snapshot of the membership set. Audience Trends aggregates this data into time-series visualizations and summary metrics.

Key Metrics

MetricDescription
Current SizeNumber of members in the audience as of the most recent evaluation
Growth RatePercentage increase in audience size over the selected time window
Shrink RatePercentage decrease in audience size over the selected time window
Net ChangeAbsolute change in member count (positive = growth, negative = shrink)
Change VelocityRate of membership changes per evaluation (adds + removes per run)
Churn RatePercentage of members who exited the audience in the selected window
New Member RatePercentage of current members who entered the audience in the selected window

Reading the Dashboard

Audience Size Over Time

The primary chart shows audience member count plotted over time. Each data point corresponds to an audience evaluation. Use this to:

  • Spot trends — Is the audience growing, shrinking, or stable?
  • Identify inflection points — When did growth accelerate or a decline begin?
  • Correlate with events — Compare audience size changes against campaign launches, product releases, or seasonal patterns.

Growth and Shrink Analysis

Below the size chart, a breakdown shows the composition of each change:

  • Members Added — New members who entered the audience since the previous evaluation
  • Members Removed — Members who exited the audience since the previous evaluation
  • Members Retained — Members who remained in the audience across evaluations

This decomposition reveals whether an audience is “churning” (high adds and removes but stable size) or genuinely growing/shrinking.

Membership Change Velocity

Velocity measures the total rate of change (adds + removes) per evaluation, regardless of direction. A high-velocity audience has a lot of membership turnover, which can impact:

  • Sync volume — High-velocity audiences generate more sync operations
  • Destination costs — More API calls to add and remove members
  • Campaign targeting stability — Members may enter and exit before a campaign reaches them

Historical Snapshots

SignalSmith stores audience evaluation snapshots in your warehouse’s audit schema. You can query these directly for custom analysis:

SELECT
  evaluated_at,
  member_count,
  members_added,
  members_removed
FROM CDP_AUDIT.AUDIENCE_EVALUATIONS
WHERE audience_id = 'aud_abc123'
ORDER BY evaluated_at DESC

Common Use Cases

Measuring Campaign Impact

Compare audience trends before and after a campaign:

  1. Navigate to Insights > Audience Trends
  2. Select the target audience (e.g., “Active Users”)
  3. Set the time window to span the campaign period
  4. Look for size changes that correlate with campaign start and end dates

If the audience grows during the campaign and stabilizes after, the campaign likely drove new qualifying behavior. If it shrinks after the campaign ends, the campaign may have temporarily inflated the segment.

Monitoring Audience Decay

Some audiences naturally shrink over time as members stop qualifying (e.g., “Purchased in the last 30 days” will lose members who haven’t purchased recently). Trends help you:

  • Set expectations — Understand the natural decay rate so you can plan for list attrition
  • Trigger re-engagement — If a high-value audience is decaying faster than expected, it may signal the need for retention campaigns
  • Adjust criteria — If an audience shrinks below a useful threshold, consider broadening the filter conditions

Identifying Anomalies

Sudden spikes or drops in audience size often indicate data issues rather than real changes:

AnomalyPossible CauseInvestigation
Sudden 50%+ dropSource data issue (table truncated, schema change)Check the source model and recent data loads
Sudden 50%+ spikeFilter condition bug or new data backfillReview the audience filter and recent model changes
Size drops to zeroSource connection failure or query errorCheck the audience evaluation logs for errors
Wild oscillationNon-deterministic query (e.g., relying on CURRENT_TIMESTAMP in joins)Review the model SQL for time-sensitive logic

Comparing Multiple Audiences

Select multiple audiences to plot their trends on the same chart. This is useful for:

  • A/B segment comparison — Compare how two variations of an audience criteria perform
  • Funnel analysis — Track audiences that represent different stages of a customer journey (e.g., “Signed Up” → “First Purchase” → “Repeat Buyer”)
  • Portfolio monitoring — See all your key audiences at a glance to spot which are growing and which are declining

API Endpoint

GET /api/v1/insights/audience-trends

Returns audience size time series and change metrics.

Query Parameters:

ParameterTypeDescription
audience_idstringFilter to a specific audience (optional, returns all if omitted)
windowstringTime window: 7d, 30d, 90d, 1y (default: 30d)
granularitystringData point granularity: run, daily, weekly (default: daily)

Response Fields:

FieldTypeDescription
audience_idstringAudience identifier
audience_namestringAudience display name
current_sizeintegerMost recent member count
data_pointsarrayTime series of { timestamp, member_count, added, removed }
growth_ratefloatPercentage growth over the window
shrink_ratefloatPercentage shrink over the window
change_velocityfloatAverage adds + removes per evaluation