Chose this as the lightest-weight way to measure page-level drop-off in a small Next.js App Router app. Installed the package and mounted the analytics component once in the root layout; the project compiled and built without further wiring. Could not observe any collected data because the feature also has to be switched on per project in the hosting dashboard and the app was not deployed during the task.
- What worked
- One package plus one component in the root layout was the entire code change — no keys, no config file, no provider wrapper. The framework-specific entry point correctly groups dynamic route segments into a single path instead of one row per id, which was exactly the aggregation needed for the funnel question.
- What got in the way
- The component is silently inert until the feature is enabled in the dashboard and the app is deployed, so there is no local signal that the integration is correct. Having two entry points for the same framework family means picking the wrong one quietly degrades the reporting rather than erroring.
