Custom Attribution Data in Marketing Mix Modeling (MMM)
How to provide ecommerce outcomes from your own attribution or MTA system for use in Sellforte modeling.
Sellforte can use ecommerce outcome data from your own internal attribution or multi-touch attribution (MTA) tracking as the business outcome signal for modeling. This lets you measure purchases and revenue using the same acquisition context your team already captures, rather than a platform-native export like GA4 or Adobe Analytics.
Custom attribution data only makes sense for ecommerce. If your business does not sell products online, there is no ecommerce outcome to track and no custom MTA data to build from, so this article covers the ecommerce case only.
Dataset definition
Ecommerce purchase outcomes from your internal attribution or MTA system, provided at the individual transaction level with the acquisition context you record, used as business outcome signals for modeling.
Because this data is fully custom, Sellforte does not prescribe a source system, extract method, or field naming. You provide the transactions in whatever way suits your setup, and the details below describe the information each row should carry rather than a fixed schema.
Grain & sample
Each row represents one ecommerce transaction per date, country, currency, and the acquisition context recorded for that transaction.
Example rows (illustrative):
| date | transaction id | country | currency | channel / source / medium | campaign | ecommerce purchases | purchase revenue |
|---|---|---|---|---|---|---|---|
| 2025-01-01 | T12345 | FI | EUR | paid search / google / cpc | spring_sale | 1 | 129.00 |
| 2025-01-01 | T12346 | FI | EUR | paid social / facebook / paid | prospecting | 1 | 89.00 |
| 2025-01-02 | T12347 | SE | SEK | direct / none | (none) | 1 | 59.00 |
Dimensions
- date
- transaction id
- country
- currency
- campaign id
- campaign
- channel / source / medium
- default channel group (or your equivalent acquisition grouping)
Metrics
- ecommerce purchases
- purchase revenue
- total revenue
- shipping amount
- tax amount
Attribution model note
Because the acquisition context in this dataset comes from your own tracking, please document which attribution model produced it (for example last-click, data-driven, or a custom rule-based model). Sellforte reads these transactions as raw outcome signals and places the results on the timeline as part of the modeling, so knowing how the upstream context was assigned helps interpret the results correctly.