The Time Decay Model (by default)
Quanticfy uses a 7 days time decay attribution model by default (shown in the table below).
In this model, hits are not weighted the same depending on when they occurred. This is relevant for most ad attribution through UTMs: ads seen (or clicked on) a week before the sale are considered less important for attribution than ads seen (or clicked on) the same day of the order. We look at the customer journey up to 7 days prior to the sale.
Sale | D0 | D1 | D2 | D3 | D4 | D5 | D6 | D7 |
coef | 1 | 0,875 | 0,75 | 0,625 | 0,5 | 0,375 | 0,25 | 0,125 |
If automatic tracking is not possible (for all media that do not provide a business manager and/or an API), like it is the case for Influence, we provide you with an interface to manually upload your costs (the Media Spending Influence). This is the only way we can collect exhaustive attribution data and provide you with Effective Media Spending and Effective ROAS.
The Flat Model
If your business relies a lot on Influence (more than 20% of your sales), we recommend you switch to another attribution model.
The Time Decay model is not suitable for Influence because it makes promo codes overweighted in the attribution, as they are always used on day 0 (the day of the sale), with no prior appearances in the customer journey. This creates a recency bias that we can eliminate by switching to a flat attribution model.
In this model, the coefficient for all days (D0 to D7) is 1, like so:
Sale | D0 | D1 | D2 | D3 | D4 | D5 | D6 | D7 |
coef | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 |
To request an attribution model change, please open a ticket (using the Intercom chatbot).
⚠ For our attribution models to work, UTMs and AD-Ids need to be properly parameterized. Learn more on our guide to do so if you need assistance.