Overview

Trend metrics in Sprinklr are used to show how a metric changes over time, with Date used as the dimension. Trend records are currently generated based on the difference between consecutive lifetime metric fetches.

For absolute metrics such as likes and views, this approach accurately represents the change in the metric over time. However, this approach is not meaningful for percentage-based metrics. For example, if an Instagram Post Skip Rate changes from 42% to 45%, a trend record of +3% represents the difference between the two values rather than the actual skip rate.

For percentage metrics, the more useful representation is the actual lifetime value recorded at each timestamp.

Updated Calculation for Percentage Metrics

For percentage metrics in scope, trend records now store the fetched lifetime value at each timestamp instead of the delta from the previous fetch.

For example, if a metric has the following values:

The trend will represent the values as 10%, 10%, and 15%, allowing the time-series visualization to show how the metric itself evolved over time.

Previously, if the value remained at 10%, the trend calculation could show 10%, 0%, 0%, because it represented the change from the previous value. The updated calculation instead preserves the actual value for each date.

Trend Records - Percentage Metrics vs Other Metrics

For percentage metrics in scope:

• Trend records store the fetched lifetime value at each timestamp.
• Trend records no longer store the difference between consecutive lifetime fetches.
• Time-series visualizations can accurately show whether the percentage metric is increasing or decreasing over time.

For all other metrics, there is no change. Trend records continue to use the delta between consecutive lifetime fetches, which remains appropriate for absolute metrics such as likes and views.

Sample Scenario

This change applies to percentage-based metrics such as Instagram Reel Skip Rate and TikTok Video Completion Rate.

For example, if a video has a completion rate of 10% on September 7 and remains at 10% on September 8, the trend will show 10%, 10%. If the value changes on a subsequent date, the new percentage value will be reflected for that date.

This provides a more meaningful representation of percentage metrics by showing their actual values over time rather than day-over-day deltas.

Limitations

• The updated trend calculation applies only to trend records generated after the DP is enabled. Existing trend records are not recalculated or updated.
• If the DP is disabled after being enabled, the trend calculation changes back from the date the DP is disabled. Trend records generated while the DP was enabled retain their existing values and are not modified retroactively.