No action until instrumentation needed. Just putting some notes here after thinking through options.
To get a better view of the full funnel, we can either:
Add more instrumentation directly to credere-backend
Add events to Fathom. BUT, Fathom is unlikely to be able to track e.g. number of emails sent, so probably makes most sense to put all instrumentation in credere-backend, instead of splitting.
Move the UUID to a query string parameter, so that the URL paths are useful for analytics.
(3) would be simplest as a "quick win", but not sure if it has any implementation challenges. (We would probably keep the existing routes in addition to the new routes, because users will have already received emails with the old routes.)
(4) is only worth exploring if (3) isn't doable or is more complicated than (4).
No action until instrumentation needed. Just putting some notes here after thinking through options.
To get a better view of the full funnel, we can either:
(3) would be simplest as a "quick win", but not sure if it has any implementation challenges. (We would probably keep the existing routes in addition to the new routes, because users will have already received emails with the old routes.)
(4) is only worth exploring if (3) isn't doable or is more complicated than (4).
FYI @sebasdocp