Applying time series to result or to contributing indicators
Summarize
Summary of Applying time series to result or to contributing indicators
In ServiceNow Performance Analytics, when using formula indicators, time series aggregations can be applied either to each contributing indicator individually or to the final formula result. This behavior is controlled by theApply time series to resultoption, found in the Other properties tab of a formula indicator record. Understanding how and when to apply this setting is crucial for accurately interpreting your metric calculations over time.
Show less
Key Features
- Apply time series to result option: Determines whether time series aggregation happens after the formula is calculated (applied to the result) or before (applied to each contributing indicator).
- This setting impacts how data is aggregated and displayed in Core UI Performance Analytics widgets, Analytics Hub, and Platform Analytics Data Visualizations.
- The option also governs the behavior of any default time series set on the indicator, but the default time series only applies in Analytics Hub and KPI Details.
- If no time series aggregation is selected in a widget or visualization, the default time series does not apply.
- Choosing a non-aggregate default time series (like simple indicator frequency) means the Apply time series to result setting has no effect.
Practical Implications
When Apply time series to result is checked:
- The formula evaluates daily or per frequency, and then the time series aggregation (e.g., 7-day running average) is applied to the formula’s final output.
- This means you get an average of the daily ratios or percentages calculated by the formula.
When the option is unchecked:
- The time series aggregation applies separately to each contributing indicator before the formula calculation.
- The formula then processes these aggregated values, producing a result based on aggregated input data rather than aggregated output.
This difference can yield substantially different results. Neither approach is incorrect; the choice depends on what you want to measure and represent.
Example Use Case
Consider a formula indicator calculating the percentage of new Priority 1 incidents daily:
Formula: (Number of new P1 incidents / Number of all new incidents) 100
- With Apply time series to result enabled and a 7-day running average time series, the formula calculates daily percentages first, then averages these percentages over 7 days.
- With the option disabled, each contributing indicator (new P1 incidents and all new incidents) is first averaged over 7 days, then the formula calculates the percentage from these averages.
Visualizing both versions in a time series widget helps you understand how this setting influences your analytics outcomes.
Additional Considerations
- The setting only takes effect if you select a real aggregate time series, not just the indicator frequency.
- Default time series apply only in Analytics Hub and KPI Details views, so manual time series selection is important in other contexts.
- Understanding these options allows precise control over how your metrics reflect trends and averages, crucial for accurate reporting and decision-making.
For a formula indicator, a time series aggregation can apply either to each indicator in the formula individually or to the formula result.
- The default time series applies only on the Analytics Hub and KPI Details. If you do not select a time series aggregation on a widget or data visualization, the default time series does not apply.
- For the setting to take effect on the Analytics Hub or KPI Details, you must choose a real aggregate, if the indicator does not have a default time series set. If the time series is just the indicator frequency (daily, weekly, and so on), theApply time series to result setting does not apply.
When Apply time series to result is checked, first the formula is evaluated and then the selected time series is applied to the final result. When Apply time series to result is not checked, each contributing indicator is evaluated and the default time series is applied to it. Then the formula is evaluated. The results between the two settings can differ significantly. Neither setting is wrong, but you have to think carefully about what you are measuring before making your choice.
Applying a time series to result compared to applying it to contributing indicators
Consider the formula indicator "% of new P1 incidents". Every day this indicator calculates the percentage of new incidents that are Priority 1 - Critical:
You decide that you want the result to display a 7-day running average by default on the Analytics Hub. In the
Other tab of the indicator record, you select the 7d running AVG default
time series. You apply the time series to the result.
In the resulting calculation, the formula is resolved for each day. Then the average of the result is taken for that day and the previous six days:
You aren't sure if you want the 7-day average of the final result or the average 7-day average of each indicator. So, you copy the previous formula indicator, with the same time series, but with Apply time series to
result unchecked. Now, the time series is applied to the Number of new incidents > Priority = 1 - Critical and Number of new incidents contributing indicators separately before
the formula is resolved:
You plot both formula indicators in a time series widget to see the difference in outcome
between the two settings. Because the default time series only applies on the Analytics Hub, you also add the 7d
running AVG time series to the widget: