Attribution

Attribution Model

Koddi uses a last-touch attribution model out of the box. In short, last-touch attribution gives the conversion credit entirely to the final ad touchpoint before the conversion event. This is the standard attribution for almost all private marketplace ad programs (and default option for most analytics and reporting tools across the ad tech landscape). It is for this reason, this is not a configurable option.

That being said, Koddi can support alternative models. Please contact your Koddi team to discuss additional options.

Attribution Types

Koddi supports multiple, configurable attribution types:

Direct: same entity clicked converts
Exposed: same entity viewed converts
Halo Direct: entity related to the entity clicked converts
Halo Exposed: entity related to the entity viewed converts

The halo grouping is also configurable – it can be any level in the campaign structure (same advertiser for example) or an attribute in the inventory feed (same brand for example).

Look Back Window

The look back window is also configurable and can be set to any numbers of days less than 30. If you have a use case for anything beyond 30 days, please contact your Koddi team to discuss additional options.

Online-to-Offline Attribution (Store Visits and In-Store Sales)

Many retailers use Koddi to measure how digital advertising influences in-store behavior. The most common approach is a CRM or loyalty-based matchback, where offline purchases or store visits are tied back to users who previously engaged with ads delivered through the Koddi platform. This allows for reliable, privacy-safe attribution without relying on device-level tracking.

How It Works

Loyalty ID is included in the Winning Ads Call (generally hashed - in the GUID or id fields depending on your use case).
Offline visit or purchase events are sent into Koddi through server-to-server ingestion in the conversion flow. Events are processed the same way as online conversions. They appear in reporting, optimization, and lift measurement alongside standard digital outcomes.
Koddi applies configurable lookback windows and deduplication rules to determine how offline actions should be attributed. Clients can use simple last-touch models or more advanced methods like weighted attribution or incrementality testing.

External Location-Based Data (Optional)

For retailers without CRM or loyalty identifiers, or those wanting to layer in foot-traffic insights, Koddi can also accept store-visit signals from external location providers. These partners supply aggregated, privacy-safe visit data that can be tied back to exposed audiences. This approach is optional and typically used when loyalty programs do not capture enough of the in-store population.

Reporting Attribution

The following performance metrics are available in reporting:

Revenue
ROAS
Transactions
Units
Exposed Revenue
Exposed ROAS
Exposed Transactions
Exposed Units
Halo Revenue
Halo ROAS
Halo Transactions
Halo Units
Halo Exposed Revenue
Halo Exposed ROAS
Halo Exposed Transactions
Halo Exposed Units