Attribution
Attribution
Attribution is a method for assigning credit for sales and conversions to touchpoints in a customer’s lifecycle throughout your website or app. Attribution is available across most Kissmetrics reports.
Attribution Models
An attribution model is the rule that figures out how to credit sales and conversions to touchpoints in conversion paths. Each model groups customers into buckets based on how they interact with your website or app.
For example, the Last Touch model in Kissmetrics assigns 100% credit to the final touchpoint/click that immediately precede sales or conversions. In contrast, the First Ever model assigns 100% credit to touchpoints that brought the customer in when we first saw them come to the site or app as an anonymous user.
Kissmetrics Attribution Models
First Ever
Will group your users according to the first property/touchpoint value they ever received (can be outside the date range).
Last Ever
Will group your users according to the last property/touchpoint value they ever received (can be outside of date range). This type of grouping is used to group users according to the value they currently or most recently have.
First in Date Range
Will group your users according to the first property value they received within the date range.
Last in Date Range
Will group your users according to the last property value they received within the date range.
Last Touch (available only in Metrics)
Will group your users according to the property value they had at the time of performing the metric’s event. This is the only attribution model that makes it possible for one user to be counted across multiple property values or touchpoints. For example: If I purchase from both San Francisco and Los Angeles, it would show that one person purchased in San Francisco and one person purchased in Los Angeles even though I’m the same person.
Attribution Model Examples
Lets walk through an example to show the output for each type of attribution model. In our example, let’s create a metric for “Total Value for Property” where our calculated property will be revenue. In our example, Bob will receive the following events and properties on the following dates:
April 16
Visited Site
Referrer: linkedin.com
May 2
Visited Site
Referrer: facebook.com
Purchased
Revenue: 50
May 5
Visited Site
Referrer: google.com
Purchased
Revenue: 100
June 3
Visited Site
Referrer: pinterest.com
Examples of Attribution Results
Here is the output for each type of grouping for the date range of May, segmented by referrer:
| First Ever | |
| Linkedin.com | 150 |
| Last Ever | |
| Pinterest.com | 150 |
| First in Date Range | |
| Facebook.com | 150 |
| Last in Date Range | |
| Google.com | 150 |
| Last Touch | |
| Facebook.com | 50 |
| Google.com | 100 |
Note: For the metrics “Conversion Rate” and “Average Time Between Event” – to be grouped in any bucket using “Last Touch”, the user must have the same value at the time of both events. Otherwise, they are excluded from calculation.