Registration Completion

Switch this journey on when the signup funnel shows a notable share of started registrations still unfinished 30 minutes later. First, registration start, the step reached and completion must be tracked on the same registration; the email address or phone number has to be collected and verified at an early step of the flow, with contact permission recorded at that point. If abandonment clusters at one step with a validation error or a confusing field, fix that step first; a reminder would only send people back into the same error.

Incremental registration completion vs control

The journey's return is how many more registrations the reminded group finishes than the control group; clicking back into the form is not a completed registration.

How it is counted: Take every registration with a reachable channel that was assigned to either group; the metric is the share completed within the same follow-up window after assignment, and the gap between the groups is the journey's effect. Reminder delivery or clicks do not change the denominator.

Also track

• Return to the unfinished stepAssigned registrations where the user comes back through the link to the step they left, as a share of all assigned; the intermediate step before completion.
• Time to completionCompare the time from registration start to completion between the two groups; if the reminder only speeds registrations up, the gap closes as the follow-up window grows.
• Result by channelRead the email and SMS branches separately but do not compare them head to head; the branch is chosen by reachability, not at random, so the two represent different users.
• Unreachable share and complaintsTrack the share of registrations ending in No reachable channel and SMS complaints; the first shows how early contact details are collected, the second whether permission was really given.

• Trigger and entryThe journey starts when a registration is still incomplete 30 minutes after it began. Reachability is checked by email first, then by SMS; users with a reachable channel are then split at random into a journey group and a control group that gets no reminder.
• Who must not enterAnyone who has completed registration or is still in the form does not enter; tune the 30-minute threshold to the flow's normal completion time. One registration is one instance: no second reminder opens for the same registration, and unverified contact details do not count as reachable.
• Timing and channelsThere is a single touch: an email taking the user to the unfinished registration step if email is deliverable, otherwise an SMS with the same content. The same reminder never goes through both channels to one person. After the message, the journey waits 24 hours and checks whether registration is complete.
• Exit and handoffIf neither email nor SMS is available, the journey ends without a message in the No reachable channel exit. At the 24-hour check, a completed registration ends in the Registration completed exit and an incomplete one ends with no further reminder; there is no handoff.
• Frequency and consentA single reminder per registration, and its link returns the user to the step they left, never to the start. Only an email address or phone number the user verified and gave permission for is used; having no reachable channel is no reason to bypass permission rules.

1. 1Registration completion level with the control group means people were coming back on their own; more clicks do not make up for it. Where there is a gap, read it alongside SMS complaints.
2. 2Your figure is tied to the 30-minute threshold, the email-then-SMS order and the 24-hour wait. Do not add what happens after registration, such as activation, to this journey's result; with no handoff, a completed registration is credited to it alone.
3. 3For a small or murky gap, enter each group's registrations and completions into the A/B test significance calculator; if it is not significant, wait for more registrations to accumulate. If the same validation error keeps recurring, investigate the signup problem rather than adding messages.
4. 4If the reminder's link lands on the start of registration instead of the unfinished step, low completion is the link's fault, not the message's; test that the deep link opens the right step before reading the result.