Model Comparison

Purpose of this page

While the rest of the dashboard is powered by the Magpie AI Attribution Model, the Model Comparison page allows you to compare different attribution methodologies side by side, such as Multi-Touch Attribution (MTA) and Last Click.

This view helps you understand how your marketing results would look under different attribution lenses, making it especially valuable for:

• Understanding model sensitivity

• Spotting discrepancies across tracking systems (e.g. Google Analytics)

• Aligning stakeholders with different attribution preferences

How it works

Attribution Logic

The dashboard uses the following attribution models in comparison:

• Magpie AI Model (default) – Hybrid attribution combining MTA, MMM, and experimental data

• Multi-Touch Attribution (MTA) – Distributes value across all touchpoints

• Last Click Attribution – Assigns 100% of conversion value to the last interaction

We strongly recommend using the Magpie AI Model as your main source of truth.
It combines multiple data sources and methods, providing a result that comes as close to reality as possible.

Each model processes the same underlying data but may lead to very different conclusions, especially at the campaign or sub-channel level.

Filters and Dimension Control

You can:

• Filter by Channel, Sub-channel, or Campaign

• Compare attribution results side by side across models for each metric in the dashboard.

• Use the Dimension Selector to switch between aggregation levels, in the table

• Switch to day, week or month level using the Date granularity option

• Compare different channels, sub channels or campaigns - based on the selected period - in the table.

Content

1. Attribution models graph

This table shows campaign or sub-channel performance across three models:

• Select the metric you want to see

• Helps answer: Would I draw different conclusions with Last Click vs MTA?

• Are the models moving consistently over time?

This graph has no option to compare different campaigns or channels. For that use the table

2. Deepdive Table

Use the dimension selector to compare campaigns, sub channel or channels for the chosen metric.

• The table works with the selected metric above the graph

• The table shows data for the entire time period that is selected

• Helps to: Spot differences between attribution models

Use this page to:

• Compare how your results differ under alternative attribution models

• See which channels or campaigns are model-dependent

• Spot potential tracking or measurement anomalies

• Align internal stakeholders who use different attribution standards

• Evaluate whether switching models would impact strategic conclusions

• Calibrate platform results with Magpie results.

Want to know how the Magpie AI model blends different sources and methods?
See the Methodology – Attribution Models section for a deeper breakdown.