Introduction to the PathFactory Recommendation Service

The PathFactory Recommendation Service is designed with your visitor in mind. Whereas conventional B2B marketing is focused on how the marketer can push their content on an audience, PathFactory’s Recommendation service flips that on its head by focusing on the unique preferences of the individual, and an understanding of content to micro-personalize content recommendations to them.

The PathFactory Recommendation Service uses the following as it’s key data inputs:

Understanding of Content – Generated by the PathFactory Content Analysis Service. Content analysis is done at both the individual content level and content cluster level to understand the key words and phrases, content summary, readability, relationships between content, hierarchy, taxonomies and topic modeling.
Visitor Profile – Firmographic data about the account of the visitor (even if the visitor is anonymous to your Marketing Automation Platform), e.g. location, industry type, employee range, revenue range, etc.
Visitor Behavior – A deep understanding of how the visitor engages with your content such as actual time spent vs. expected time spent, scroll depth on the content and click steam data with Website Journey Tracking.

The PathFactory Recommendation Service is initially designed for use by Website Tools Components (Guide & Concierge) within Website Tools, however it will be expanded to include The Website Tools Recommendation API, and later to power other modules within the PathFactory Platform.

Website Tools provides two types of algorithms that provide different types of content recommendations

Recommended For You
Recommendations are tailored to specific visitors based on their behaviour, how content is related, and demographic information.

Related Content
Recommendations are related to the content that the visitor is currently viewing. These recommendations are based on related content.

Recently Visited
Recommendations are based on content that a visitor has recently visited. These recommendations are based on visitor behavior and allow visitors to pick up where they left off on previously viewed content. Content shown in this carousel is what the visitor has visited in the past 30 minutes (subject to change). The order of the recommendations is from most recent to least recent.

Trending
Recommendations are based on the content engagement of all visitors to the web property in the past 30 days and can be shown for all visitors to the website (Website) or filtered to visitors that share an attribute in common with the visitor (Region, Industry, Account). Engagement is defined not simply as the content that is clicked on or accessed the most but also factors in how deeply the content was engaged with (time spent and scroll depth).

The trending parameters are:

• Website – Based on all visitor engagement and behaviour on your website property.
• Region –  Based on what is trending amongst visitors to the website who are from the same state or province as the visitor who is browsing your website property.
• Industry – Based on what is trending amongst a visitor’s peers within the same industry.
• Account – Based on what is trending amongst a visitor’s colleagues.