Getting started
Required setup
Optional setupTutorials and examples
Onboarding guides
Tips and tricks
P&L Report
Target metricsMore information
Attribution models
Bayesian statistics
Frequently asked questions
Metrics overview
User journey tracking
Product level metricsData UpdatesAttribution models
Our attribution models are based on the user journeys that are tracked through our tracking pixel integration. Once tracking is enabled, user journeys are tracked considering user consent is provided. If the user did not provide consent, or tracking is disabled, we will retrieve the last landing page/touchpoint through Shopify's API.
The big picture
There is no right or wrong attribution model. Each model has it’s benefits and downsides. Looking at the output of different models will paint a picture that helps understand the contribution of each channel and campaign and guide your decision-making.
The Admetrics Data Studio offers five attribution models which you can easily switch and compare.
Attribution model overview
First touch attribution
The first touchpoint within the attribution window that is not direct gets 100% of orders and revenue attributed.
What is a lookback window?
A lookback window describes the number of days touchpoints are considered for attribution prior to the conversion. Our attribution models can be customized to suit your needs, including 1 day, 2 days, 7 days, 30 days, and Max (infinite).
Session timeout
User sessions do last a maximum of 30 minutes. 30 minutes after the first event that has been logged for a specific session, there will be a new session created. Each session results in a visit.
Last touch attribution
The last touchpoint within the attribution window that is not direct and not organic gets 100% of orders and revenue attributed.
Last platform touch attribution
The last touchpoint of each paid platform within the attribution window gets 100% of the orders and revenue attributed. This model will inflate and duplicate orders across platforms. The Direct channel is ignored by this model and won’t get any credit.
This model is supposed to address issues that exist in other attribution models where touchpoints for platforms will not be fully credited because the users also clicked ads on other platforms. So this model is an excellent choice to learn about the relative performance of activities on a specific platform.
This model can be used to compare results against media partner self-attributed numbers for platform specific performance optimisation. For cross channel performance evaluation you should rather use a fractional multi-touch model, like linear or position based attribution.
Linear attribution
Each touchpoint within the attribution window that is not direct and not brand search gets an even share of orders and revenues attributed. Example: 3 touchpoints each get 33.3% of the order and revenue.
Paid brand search channels are mostly excluded from attribution in this model. This is configured through the Channel group setting in the channel configuration (”Paid Branded Search”).
Position-based attribution
The first and the last touchpoint that are not direct and not brand search within the attribution window get 40% of the orders and revenue attributed, all other non-direct & non-brand-search touchpoints get an even amount of the remaining 20%. Example: 4 touchpoints - the first and the last get 40%, the other two each 10% attributed.
Paid brand search channels are mostly excluded from attribution in this model. This is configured through the Channel group setting in the channel configuration (”Paid Branded Search”).
Inflated attribution
Any touchpoint within the attribution window gets 100% of the orders and revenue attributed. This model will inflate and duplicate orders even within platforms.
This model is supposed to address issues that exist in other attribution models where touchpoints for platforms will not be fully credited because customers also clicked on other ads on the same platform. So this model is an excellent choice to learn about the relative performance of activities even within a specific platform or channel
For cross channel performance evaluation you should rather use a fractional multi-touch model, like linear or position based attribution.
Media partner self-attribution
The following chart visualizes the limited view of a user's journey that is available to the actual media partners.
