Predictive Segments: Likelihood to Purchase

In Digital Marketing, it is becoming more expensive to acquire new customers whereas their lifetime values are getting lower. Providing a valuable segment which will bring more conversions and improve ROAS (return on ad spend), the Likelihood to Purchase (LTP) algorithm aims to help companies and agencies use their digital marketing budgets more effectively.

Determining what behaviors lead to a purchase (e.g. multiple visits on the same item, device type, cart page) and giving a coefficient to these behaviors in its model, this algorithm learns from the buyer behavior and predicts the visitors that are more likely to make a purchase.

The LTP algorithm works in the following flow:

• The algorithm learns from the behaviors of the users who complete a purchase and determines what kind of behaviors lead to a purchase.
• It gives a coefficient to these behaviors through its model.
• Each user is assigned a score regardless of whether the user is logged in or anonymous.
• Whenever a user comes to the website, their score changes in real-time based on their behavior on the website.
• Once their score passes the threshold that is determined automatically by the model, this user is expected to make a purchase within the next 7 days.

You can use high likelihood to purchase segments on ad platforms such as Google AdWords, Facebook Ads and other digital marketing channels to convert more clicks into purchases.