Getting Started with Brightcove Recommendations

What is Brightcove Recommendations?

Brightcove Recommendations is a native recommendation engine, API, and player plugin that automatically suggests the next best videos to watch. It is fully controlled by your business and editorial rules and powered by metadata, real-time signals, and AI.

Key capabilities:

• AI-powered relevance — Recommendations are driven by semantic understanding of your content, enhanced by freshness and real-time engagement signals.
• Editorial and business controls — Define guardrails using tags, custom fields, duration limits, and restrictions to ensure recommendations stay on-brand.
• Promoted content support — Highlight priority or sponsored videos using goals, frequency, and placement without breaking relevance.
• Native player delivery — Recommendations render directly in the Brightcove Web Player with no additional frontend integration required.

Brightcove Recommendations helps you organically extend viewing sessions, increase views per visit, and surface more of your catalog without manual curation or personalization.

What it is NOT

• It is not personalized per viewer. Viewer IDs and watch history are not used.
• It does not recommend Live assets.
• It does not re-evaluate context after every recommendation. All recommendation logic is anchored to the initial video the viewer selected.

Prerequisites

• Player version 7.41.6 or later is required.
• If you were previously using the Iris.TV plugin, it must be removed before enabling Brightcove Recommendations to avoid overlay conflicts.

Enabling Recommendations

Follow these high-level steps to get started:

1. Upgrade your player to version 7.41.6 or later in the Players module.
2. Enable Recommendations in the Player module — Navigate to your player's settings, then Additional Features > Recommendation Engine and toggle Enable Recommendation.
3. Configure rules in the Admin module (recommended) — Navigate to Admin > Recommendation Engine Settings to define eligible content rules, restrictions, and promoted content.

Indexing & Readiness

When Brightcove Recommendations is enabled, the system indexes your existing video catalog so the recommendation algorithms can analyze and compare your content. Each video takes approximately 50 ms to process. Indexing happens automatically in the background after enablement.

Example: 1,000-video catalog

• Estimated embedding time: ~50 seconds
• Recommendations are usually available within a few minutes
• You may briefly see fewer recommendations or lower relevance during the first minutes

Example: 100,000-video catalog

• Estimated embedding time: ~83 minutes
• Indexing may take over an hour depending on system load
• Recommendations will appear progressively as indexing completes

Supported Delivery Methods

There are two ways to deliver Brightcove Recommendations to your audience:

Measuring Impact

Brightcove Recommendations adds a Recommendation Lift metric to the Analytics module. This metric tracks video views that were generated as a direct result of recommendations.

You can find recommendation analytics in two places:

• Recommendations Analytics section — Shows total lift broken down by algorithm (Related Content, Trending Now, Recently Added), top recommended videos, and top sources of recommendations.
• Performance module — Includes a Recommendation Lift column in the per-video performance table, allowing you to compare organic views against recommendation-driven views.

Limitations

• No viewer-level personalization — Viewer IDs and watch history are not used. Personalized recommendations are planned for a future release.
• No Live asset recommendations — Live channels and events are not eligible for recommendations.
• Anchored context — Recommendation logic remains anchored to the initial video and does not re-evaluate after every hop.