Derive data impacted by events

If you have data impacted by an event, you can use calculated metrics to derive estimated values for the duration of the event. For example, if you had an event that caused a 25% drop in data, you can use that as a multiplier in a calculated metric.

These steps work best when you understand the impact of an event, both from a segmentation and date comparison perspective. Make sure to follow Compare dates impacted by an event to previous ranges and Exclude specific dates in analysis before following this page.

NOTE
This approach is an estimate based on a specific set of inputs and date ranges. It will not be a comprehensive solution for all use cases or slices of data. In addition, this approach requires that the affected date range have at least 1 hit to calculate from.

To create an estimated calculated metric for the affected time period:

  1. Create two segments for ‘Affected days’ and ‘Exclude affected days’, as outlined under Exclude specific dates in analysis.

  2. Navigate to Components > Calculated metrics.

  3. Click Add.

  4. Drag both of the above segments to the definition canvas. Change the operator between them to a + to sum them.

  5. Add the desired metric inside both segments. For example, you could use the ‘Visits’ metric.

    Segment builder

  6. Click Add in the upper right of the ‘Affected days’ container, then click Static number. Set the static number to the percent that you want to offset your data, as outlined under Compare dates impacted by an event to previous ranges. In this example, the offset is 25%, or 1.25.

    Static number

  7. Apply the “corrected” metric side-by-side in a trended freeform table. All days outside the event reflect their normal metric count, while all affected days use the multiplier offset.

    Corrected metric

  8. View the data in a line visualization to see the effect of your corrected metric.

    Corrected line

recommendation-more-help
df89a595-bae4-48f4-9606-e55c9666ee49