Loyalty Tier Change

Reach for this journey when the program report shows upgraded members never touching their new benefits, and downgraded members learning about the drop from support rather than from you. Confirm that the tier change event carries its direction, that the benefits the new tier adds can be told apart from those the old tier already had, that the distance to the higher-tier threshold can be computed, and that email and push consent exist. If tiers are moving in bulk because of a calculation error or a rule change, fix that first; announcing a change that is wrong misleads the member instead of informing them.

Incremental new-benefit use vs control

For the upgrade branch, the value is how many more members use a new-tier benefit after the email and push than in the control group; a message open, or use of a benefit the old tier already had, does not count.

How it is counted: Among all members assigned at entry to either group who fall on the upgrade branch, the share who use one of the benefits the new tier added within the same follow-up window from assignment; the difference between the two groups is the journey's effect on upgrades. The downgrade branch is read the same way with its own outcome, moving closer to the threshold, and delivery does not change the denominator.

Also track

• Downgrade threshold progressOf the members assigned on the downgrade branch, the share who move closer to the higher-tier threshold 7 to 14 days later, compared with control; this is the result of explaining the way back.
• Use after the reminderUpgraded members with no benefit use on day 7 who use one after the push, as a share of that group; it shows whether the email alone was enough.
• Direction mixTrack how many entries are upgrades and how many are downgrades; the two branches differ in size, so read them apart instead of pooling them.
• Unsubscribes and support contactsWatch email unsubscribes after the upgrade email, push opt-outs after the downgrade push, and support contacts asking why the tier dropped.

• Trigger and entryEach change in a member's loyalty tier opens the journey, and the first step checks the direction. At entry, before the direction is read, randomly assign changes to the journey group or to a control group that gets no message from this flow.
• Who must not enterA single reward being credited is not a tier change and belongs to Reward Confirmation; a new membership's starting tier is Loyalty Program Activation's job, and anniversaries are Birthday & Milestone's. Members whose tier change and new benefit set cannot be verified stay out; the scope is one change.
• Timing and channelsUpgrades get an email on the new tier and only the benefits it adds, then one reminder push if none is used within 7 days. Downgrades get a push on the path back to the higher tier, with threshold progress checked 7 to 14 days later. Never present an old benefit as newly unlocked.
• Exit and handoffOn the upgrade side, the journey ends when a new benefit is used, or closes after the reminder push if none is. On the downgrade side, a member moving closer to the threshold leaves into normal loyalty communication, and one who is not ends the journey. None of these exits hands off to another journey.
• Frequency and consentAn upgraded member gets at most one email and one push, a downgraded member only one push, each channel used only where its consent exists. The path-back push is sent only when the program really offers a way back to the higher tier; an extra discount is not part of this flow.

1. 1If the new-benefit gap on the upgrade branch does not beat control, members were finding the benefit without a message; email opens or use carried over from the old tier do not change that. Read the downgrade branch on its own, by the gap in threshold progress.
2. 2These results hold for the direction split, the 7 day wait on upgrades and the 7 to 14 day wait on downgrades. Use of a reward credited at the same time is credited to Reward Confirmation, and a new membership's starting tier never opens this journey. Progress made after a member leaves into normal loyalty communication is not this flow's result.
3. 3Downgrade entries may be fewer than upgrades; for each branch, enter the two groups' members and outcomes into the A/B test significance calculator separately. If the gap is not significant, wait for more changes to accumulate, and never pool the two branches into one result.
4. 4Where the downgrade push went out but the program offers no real way back, threshold progress cannot be measured; close that branch and do not read its result. If the upgrade email described an old benefit as new, the usage data measures rights already in use, not the new ones.