Abstract: Provided are a system and method for obtaining, according to current and historical URL placement(s) on search engine results pages (SERPs) for identified keywords, predicted URL performance in SERP positioning. Such performance can, using URL titles and snippets appearing on SERPS, be assessed for URLs filtered to meet targeted (i.e., non-branded, informational, transactional) classifications. When a machine learning model receives input, including, for example, (a) historical trajectories of SERP positions for filtered URLs matched for identified keywords, (b) URL classification as informational or transactional, and (c) URL work type categorization (e.g. optimization, new Content, or no action), positioning for an analyzed URL on one or more SERPs returned for query according to one or more identified keywords can be predicted for a predetermined time horizon.
Type:
Grant
Filed:
December 24, 2024
Date of Patent:
July 8, 2025
Assignee:
Terrakeet LLC
Inventors:
Jennifer Brussow, Patrick Danial, Christopher Pasquarella