Post-Trial Recovery

This journey is worth building when many trials end without a paid plan and the current recovery email sends everyone the same sales pitch. Trial end, paid conversion and the core value event must be tracked on the same account; you also need to read which setup step each account left unfinished, and to hold the email address and subscription. If trials end without the core value ever being reached, the problem sits inside the trial's first setup experience; one email after the end will not make up for it.

Incremental paid conversion rate vs control

What you are measuring is how many more ended trials become paid than in the control group; an opened email or a visit to the plan page is not a conversion.

How it is counted: The denominator is every ended trial assigned at entry to the journey group and the control group; the numerator is the accounts that genuinely move to a paid plan within the same follow-up window, and the gap between the groups is the journey's effect. Non-openers stay in the denominator, and delivery or click status does not change it.

Also track

• Conversion by value reachedRead the journey-versus-control gap separately for accounts that reached core value and those that did not; the two branches' emails do different jobs, and one failing should not hide the other's result.
• Unfinished setup step completedIn the branch that never reached value, the share of accounts that complete the missing setup step or reach out through the support route after the email; the intermediate step before a paid plan.
• Unsubscribes and complaintsWatch unsubscribes and spam complaints after the recovery email; a user whose trial ended may have lost interest, and a rise here means the copy or the send itself needs rethinking.

• Trigger and entryEntry happens when a trial ends with no paid conversion. Randomly split ended trials at entry into the journey group and a control group that gets no recovery email; make the split before the core value branch so both groups carry the same mix.
• Who must not enterAccounts whose trial is still running or was extended, and accounts that paid before the end, stay out. Each instance follows one expired trial: no second recovery attempt opens for the same trial, and an account that already left this flow does not re-enter.
• Timing and channelsAt entry the journey checks whether the core value event was completed. Accounts that reached it get an email reflecting the features they used and the relevant plan; the rest get an email showing the unfinished setup step and the support route. Then the journey waits 7 days and checks for a paid plan.
• Exit and handoffIf the account has converted by the end of the 7-day wait, the journey ends at Paid conversion completed. If not, recovery communication stops and the journey closes; there is no handoff and no second attempt.
• Frequency and consentOne email per ended trial, sent only to accounts with an active email subscription. The only gate on content is whether the core value event was completed; trial extensions, price changes and unlocking extra features are separate offer decisions, not part of this journey.

1. 1When the journey group's paid conversion rate fails to beat the control group, the email only pulled forward accounts that would have paid anyway; opens and clicks do not make up for that. Read the conversion gap together with the unsubscribe count and decide on that.
2. 2Everything above holds for these two email versions, a single touch and the 7-day check. This flow carries no offer, so do not report it as a discount campaign, and when it ran in the same period as a trial extension or a price change, do not credit the whole gap to the email.
3. 3When the conversion gap is small or its direction unclear, enter each group's ended trials and paid conversions into the A/B test significance calculator; if the difference is not significant, hold off on declaring success, let more ended trials accumulate and repeat the read separately for each value branch.
4. 4The email for the branch that reached value must describe only features the account actually used; a message praising a feature that was never opened is in the wrong branch, so fix the branching rule before reading the result. Do not try to widen the gap with a second recovery email; stop after the one recovery window.