Attribution Anomaly: The Case of the Overlapping Campaigns
An e-commerce company runs two campaigns simultaneously: a display campaign and an email campaign. They use Adobe Analytics with a 30-day visit expiration for the campaign variable (v0) and a 30-day visitor expiration. The marketing team notices that the email campaign is getting credit for most conversions, even though the display campaign is driving a lot of traffic. They suspect that the email campaign is being over-credited because it is often the last touchpoint before purchase. The analytics team reviews the data and sees that many users who clicked the display ad later opened an email and then purchased. The company wants to understand the true contribution of each campaign and optimize their budget.
1. Explain why the email campaign appears to get more credit and how Adobe Analytics attribution works in this scenario.
Read the worked answer
In Adobe Analytics, the campaign variable (v0) uses a visit-based expiration by default (30 days). When a user clicks a display ad, v0 is set to the display campaign. If they later open an email and click through, v0 is overwritten with the email campaign (since it's a new visit and the expiration allows overwrite). On purchase, the conversion is attributed to the last campaign value in v0, which is the email campaign. This is 'last touch' attribution within the visit expiration window. Additionally, if the user opens the email in the same visit as the purchase, the email campaign gets credit. The display campaign only gets credit if it was the last campaign before purchase or if the user purchased in the same visit as the display click without later email click. So the email campaign appears to get more credit due to last-touch attribution and the overwrite behavior.
2. What methods can the company use to analyze the true contribution of each campaign, and what recommendations would you make?
Read the worked answer
To analyze true contribution: 1) Use Adobe Analytics' Marketing Channels, which have configurable attribution (first touch, last touch, linear) and expiration. Set up marketing channels for display and email with appropriate rules. 2) Use the 'Attribution' feature in Analysis Workspace (Attribution IQ) to apply different models (e.g., linear, participation, same touch) to the campaign dimension. 3) Use the 'Campaign' dimension with 'Participation' metric to see which campaigns were involved in any part of the conversion path. 4) Implement a multi-touch attribution solution like Adobe's Customer Journey Analytics or use a data warehouse to analyze paths. Recommendations: 1) Switch to a multi-touch attribution model that gives credit to all touchpoints (e.g., linear or U-shaped). 2) Use Marketing Channels with a first-touch model for acquisition and last-touch for conversion, and compare. 3) Ensure campaign expiration is set appropriately (e.g., 30 days) but consider using visitor expiration for longer consideration cycles. 4) Run a holdout test or incrementality test to measure true lift. 5) Allocate budget based on multi-touch attribution results, not just last touch.