AI-Powered Content Recommendations

Overview

Hearsay Social’s AI-powered content recommendation engine provides personalized and automatic Post Library content suggestions. This feature uses data related to advisor/agent interests and behavior, such as keyword searches and post dismissals, as well as audience engagement to suggest relevant content for advisors/agents to post on their social media channels and improve their connections with their clients, prospects, and leads. 

Our content recommendation engine constantly learns from user data to provide recommendations that are customized to specific interests and contexts, similar to popular apps like Spotify and Netflix. The more advisors/agents use Hearsay Social, the more content recommendations are tailored to their preferences and their audience’s interests, making it easier to find  relevant and meaningful content to drive business. 

Recommended posts can be seen on the Recommended tab in the advisor/agent Post Library, which supports both 3rd party and admin-created suggested posts, depending on your organization’s configuration. The regular Post Library still exists in the Browse tab.

There are six sections of content in the Recommended tab of the Post Library, as follows:

  • Highlights at <your organization>: Content recommended by content creators/admins using the Pinned Posts feature. For this section to appear, at least one pinned post must be available to the workspace.
  • Recommended for your audience: Posts about topics that the advisor/agent’s audience is engaging with the most, as determined by our in-house recommender engine. For this section to appear, the following must be true:
    • There are posts in English available for the workspace
    • There are link type posts available for the workspace.
    • There is at least one audience engagement on posts published from the Post Library by the workspace.
  • Based on your interests: Content based on the advisor/agent search and publishing history, as determined by our in-house recommender engine. For this section to appear, the following must be true:
    • There are posts in English available for the workspace.
    • There are link type posts available for the workspace.
    • The workspace has accomplished at least five activities, which can be either publishing a post or performing a search in the Post Library.
  • Most recent: The newest content added by content creators/admins. For this section to appear, at least one post must be available in the Post Library for the workspace. 
  • Most engaging: Content receiving a lot of audience engagement when posted by the advisor’s/agent’s peers. For this section to appear, at least one post must be available in the Post Library for the workspace.
    Popular with your peers: Content being published a lot by advisors/agent’s peers in their organization. For this section to appear, at least one post must be available in the Post Library for the workspace. 

How does it work

New recommendations are processed in batches and sent to advisor/agent workspaces once per day. Posts can only appear if they are released or pre-released and active, and not yet published by the advisor/agent. Campaign-only posts, or posts that are part of a campaign the advisor/agent is subscribed to are not included in the recommendations.

Our content recommendation engine uses AI in two stages to find and recommend content for advisors/agents to use. We first use AI to topically categorize the articles linked in the advisor/agent Post Library, and we then use AI to make recommendations based on interest profiles created for advisor/agents’ social media audience.

AI-Powered Categorization

Our content recommendation engine is trained to analyze articles from trusted news portals to categorize articles linked in advisor/agents’ Post Library. This categorization process analyzes the title and text of articles to classify them under the following topics:

  1. Finance (& economy, tax)
  2. Insurance
  3. Investment
  4. Property (& mortgage)
  5. Retirement
  6. Sport (& health)
  7. Career (& education)
  8. Relationship
  9. Lifestyle (home, travel, food)
  10. Technology
  11. Environment

Note: The categorization process excludes articles about religious or national holidays.

AI-Powered Suggestion

After categorizing the content in your Posts Library, our content recommendation engine suggests content for advisors/agents based on interest profiles created for them as well as their social media audience. This interest profile is created by analyzing the advisor/agent’s social media publications, keyword searches, post dismissals, and campaign subscriptions as well as their audience’s engagements with them on social media, such as likes, shares, and comments. Our AI matches these interest profiles with the content categories to suggest posts that will foster engagement on social media.

Limitations

Our content recommendation engine was only trained to categorize and recommend articles written in English.

Our categorization process only analyzes the title and text of articles. It does not analyze videos, images, PDFs, or any other kind of interactive audio-visual component that may also accompany the text.

New suggestions are provided once per day. Creating and updating interest profiles as well as the categorization of articles happens overnight in your local time zone. 

Advisors/agents have to publish at least 5 posts with links to textual content for recommendations to be provided. If they have not published a sufficient number of posts, they will be shown a message encouraging them to post more.

 

AI-Powered content recommendations

AI_Recommended.png

 

Was this article helpful?
0 out of 1 found this helpful

We're sorry to hear that.

Articles in this section

See more