Highlighted

Marketing Channels & Using Segments to Define Historical Channel Data

cvonholle

29-10-2018

My organization has been operating with adobe analytics for a long time, however not all of the channel definitions were set as well as they could have been. However, there have been other definitions (like cp tracking codes) that were tracked to help define channels, campaigns, etc., on an isolated report basis.

Because updates to marketing channels do not apply retroactively, I am creating segments that can be applied to our historical data that will act as a historical channel definition. To define these segments, I am applying channel-specific dimensions such as cp codes, referring domain definitions, or previously accurate channel definitions to visits that fall in line with our newly specified channel criteria.

To ensure a strong level of accuracy, I am referencing data collected post channel definition updates and comparing marketing channel overview data against my segmented data. The data I am seeing shows significant discrepancies between the marketing channel overview report and my applied segment data for the same days in Ad Hoc analysis. In fact, I am seeing discrepancies in the numbers from the marketing channel overview and last touch channel reports and Ad Hoc Analysis reports using the "marketing channel = X" dimensions, which should produce identical data.

I have reviewed and adjusted the hit, visit, visitor containers to ensure the data is correctly being applied to and/or excluded from each segment, and often see that most metrics correctly align at least with a few columns (like unique visitors, visits, or page views) but often have inflated revenue numbers, and order numbers that are all over the place.

Has anyone else ever run into this historical channel segment scenario or any other metric discrepancy between pre-existing reports and those created in Ad Hoc or Analysis Workspace?