MCP Campaign Reporting – Segments vs Goal Completions

Business Challenge

If I want to look back over the life of a campaign to answer “how many people clicked on an experience and purchased in the last 90 days with an attribution window of 1 week” I get completely different Segment vs Goal results.

MCP Segments and Goal Completions calculate results in different ways:

  • Segments do not use attribution windows, or sequencing logic to say a user “clicked an experience and then made a purchase”
  • Whereas a Goal Completion is counted when a user enters the Goal Segment during the Web Statistics UI date range and attribution window selected
  • Segments in the User Segments UI are composed of unique individuals; they are only counted in a Segment once.
  • However, users can leave and enter Segments set up as Goals multiple times, resulting in multiple Goal Completions, but in this scenario Goal Completions do not reflect unique individuals.
  • If a user is already in a Goal Segment before a campaign is published, they will not count as a Goal Completion until they leave and re-enter the Goal Segment

This makes it challenging for marketers to report accurately on campaign influence on revenue in MCP.

To count the number of purchases made during a campaign period, this setup has been the most accurate when comparing Goal Completions to Orders in the Web Statistics UI:

  • Create a Segment and Goal for users who have “Purchased in the past day”
  • Choose this Goal and select a timeframe of 90 days in the Web Statistics UI

Here are a couple of scenarios that will occur using this approach:

  • User A makes 3 purchases in one day – they will only be counted once as a Goal Completion. This is because the user entered the Goal Segment after completing the first purchase
  • User B makes a purchase 3 times on different days – they will be counted 3 times in the Goal Completions metric within Web Statistics. This is because the user left and re-entered the Goal Segment multiple times

To count the number of people who clicked and purchased during a campaign period, create a Segment and Goal for “Purchased in the past 90 days”. This Goal Completion would only count users once over 90 days.

Goal Completions vs Google Analytics Conversions

Goals in MCP are unlikely to match conversion rates in Google Analytics exactly because these two platforms have very different tracking methods. Therefore it is not an expectation that metrics from these two platforms will match.

Most organsations use third-party analytics software to track campaign effectiveness, in addition to using Personalization Campaign Statistics. In such cases, it is possible that Personalization campaign statistics differ from statistics gathered via third-party software. This difference is mainly because tracking parameters and their definitions on Personalization can vary from how they’re defined, handled, and tracked in third-party systems.