This article describes how to monitor the health of your conversion measurement and implementation.
Consider these articles for more information on events:
Understand the health of your events
There are two ways to check event health without leaving the Events page:
Event status on the events list: Each event shows a atatus of Active, Inactive, or Sudden change, along with a 7-day trend sparkline, 7-day fire count, and the number of connections sending that event.
Current connections: Below the events list, a summary breaks out each connection: Conversions API, Google Tag Manager, JavaScript pixel, and partner integrations, and shows how many of its events are active, inactive, or flagged with a sudden change.
Verifying individual event health
Click into any event to open its detail view. This is where the new diagnostics live:
Days toggle: Switch between a 7-day and 30-day view of event volume.
Volume trend chart: Plots daily event fires over the selected window.
Sudden change alert: When volume shifts sharply from one day to the next, a callout flags the date and notes that it “could indicate a tracking issue or a natural change in user behavior.”
Event parameter coverage: Broken out by connection (if you use more than one connection, each will be available in their own tab), this table lists each parameter: email, phone, client IP address, client user agent, and advertising IDs like Roku, Google, Apple, and generic with its coverage percentage. Parameters below full coverage get an “Increase parameter coverage” action; parameters sending 0% get a “Start sending parameters” action.
Note – Users with administrator and campaign manager permissions will be able to mark as expected. If they click it, it will dismiss the alert.
Settings & diagnostics page
The Settings & diagnostics page is organized into three tabs based on how you send events: JavaScript pixel, Conversions API, and Partners. Each showing:
Recommendations: Callouts to improve your pixel and event implementation, or confirmation that everything is running smoothly and no action is needed.
Event implementation: Counts of inactive events and events with a sudden change, each with a “Show event” link.
Base pixel code: Confirms the last time your base pixel code fired (timestamp in UTC), with a link to view your base code.
How does it work?
Event-level health checks use the event-tracking code you've placed on your website or app to monitor fluctuations in event volume. Daily event-connection volume is compared to a trailing 7-day average to catch sudden changes broken out by connection type.
Who should use it?
New & existing advertisers: Quickly identify and fix setup issues, allowing accurate data collection from the start. Regardless of campaign size or technical expertise, everyone benefits from monitoring event-tracking health.
Technical teams: Get detailed information for troubleshooting code-related problems, including which parameters are missing on which connection.
Marketing and analytics teams: Monitor overall event tracking health and catch volume anomalies before they affect reporting.
Why is it important?
Spot issues: Easily identify, prevent, and correct issues with events.
Correct errors: Early detection minimizes data gaps and ensures consistent performance.
Maximize performance and reporting: Reliable event data powers optimization, attribution, and reporting for current and future campaigns.
Optimize event setup: Refining your event tracking code helps you capture all relevant user interactions.
Catch volume anomalies fast: Sudden change alerts flag tracking issues the day they happen, not weeks later in a report.
Improve match rates: Parameter coverage shows exactly which identifiers are missing per connection, so you know what to fix to improve Roku's ability to match a user for conversion tracking and audience targeting.