Performance & Metrics

Froomle provides detailed metrics to help you understand the performance of your recommendations and the behavior of your users.

Metrics overview

Metrics are aggregated in time slices and can be filtered by module, page type, or channel. Key groups include:

  • General metrics (pageviews, item interactions, purchases).

  • Integration metrics (requests, recommendation_sets, clicks, impressions).

  • Attribution metrics (attributed revenue, attributed purchases).

For the full metrics schema and query options, see Metrics API Reference.

Metric definitions (overview)

The dashboard and API expose the following common metrics:

General metrics

Name

Description

pageviews

Aggregates view events across all known pages.

item_interactions

Aggregates item_interaction events.

add_to_carts

Aggregates add_to_cart events across all pages and page types.

purchased_items

Aggregates purchases of individual items.

batch_opens

Aggregates batch open events across all integration types.

revenue

Aggregates the total price of purchases.

Integration metrics

Name

Description

clicks

Aggregates click_on_recommendation events across all integration types.

requests

Aggregates all requests across all integration types.

recommendation_sets

Recommendations generated by Froomle, distinct by request_id.

impressions_sets

Impressions across all integration types, distinct by request_id.

batch_clicks

Aggregates batch_click_on_recommendation events.

batch_requests

Aggregates batch requests across all integration types.

batch_recommendation_sets

Batch recommendations generated by Froomle, distinct by request_id.

batch_impressions_sets

Batch impressions across all integration types, distinct by request_id.

Attribution metrics

Name

Description

attributed_purchased_items

Purchases of items recommended by Froomle within the attribution window.

attributed_revenue

Revenue from purchases attributed to Froomle recommendations.

attributed_item_interactions

Item interactions attributed to Froomle recommendations.

Typical news publisher metrics

These are common metrics used by news publishers when evaluating recommendation performance:

  • Item CTR - Clicks divided by impressions at the individual item level.

  • List CTR - Clicks divided by impressions for the full recommendation list (module level).

  • Page CTR - Clicks divided by pageviews where a recommendation module was shown.

  • Recommendation coverage - Share of pageviews where a recommendation module rendered successfully.

  • Clicks per session - How many recommendation clicks occur within a user session.

  • Recirculation rate - Share of sessions with at least one recommendation click.

  • Article recirculation rate - Percentage of sessions that continue after viewing an article.

  • Article depth - Average number of articles read per session.

  • Session duration - Measured as pages per session, article-detail pages per session, and total time.

  • Subscriber conversion rate - Share of users who convert after interacting with recommendations.

  • Paid vs. free mix - Proportion of recommended clicks on paid vs. free articles.

  • Freshness distribution - Share of clicks on items within a recency window.

  • Retention - Percentage of users who return per day.

Related topics