Analyze data impacted by events

Sometimes an event can impact data quality in your organization. Examples include:

  • A bot sending outlier data, such as millions of dollars in revenue
  • Your organization pushed an update to your website that negatively impacted your Analytics implementation
  • Other issues that affect data quality or completeness

If your site experienced any kind of data quality issue, you might want to exclude it from reporting to prevent making business decisions on it. Use these sections to gauge the impact of the event on your data, and determine how you want to proceed.

Determine the cause of an event

If you are not sure why you see a spike or drop in data, see Troubleshoot spikes/drops in data.

Analyze and exclude data using segmentation

Adobe Analytics offers a simple and robust way to focus on or exclude data using segmentation. You can use date range dimensions within segments to filter out or focus on those specific dates. See Exclude specific dates in analysis.

Compare an event to previous date ranges

If you want to learn more about how much impact an event made on your data over time, you can use date comparison in Analysis Workspace. This feature lets you compare data day-by-day, week-by-week, or month-by-month to see how it compares with previous ranges. You can then use this comparison to determine how much an event affects trends. See Compare dates impacted by an event to previous ranges.

Derive data using calculated metrics

Once you create segments and use date comparison, you can combine both of these concepts to correct trended data using calculated metrics. Include the segments within a calculated metric, then multiply the affected days by the offset found when comparing dates. See Derive data impacted by events.

Communicate impact to users in your organization

Once you are prepared with how you intend to handle an event, you can communicate to users in your organization. Adobe offers several places within Analytics where you can place text to communicate to users what happened and what components they can use.

Video

This video walks through each of the above steps.

  • 0:27: Exclude data using segmentation
  • 2:55: Compare an event to previous ranges
  • 8:42: Derive data using calculated metrics
  • 11:46: Communicate impact to users
recommendation-more-help
df89a595-bae4-48f4-9606-e55c9666ee49