Rule examples
Standing rules for automated customer actions.
Each rule is a policy that runs on a schedule: the conditions a customer must meet, the exclusions that remove them, and the audience it feeds. The campaign becomes a standing rule, not a recurring spreadsheet task.
Two examples
One rule from order history. One that adds a signal from text.
Both are illustrations of how a rule reads, not client policies.
- Completed purchases
- 2 or more
- Usual reorder interval
- 30 days
- Days past expected reorder
- 10
- Historical revenue
- Over $150
- Email consent
- Still subscribed
- Recent contact
- None
- Exclusionscancellation, refund, other suppression
- None
Feeds the client's Klaviyo workflow. Customers who meet every condition are eligible today.
- Quiz signalextracted from a quiz response, source text retained
- Goal: sleep support
- Product A purchases
- 2
- Reorder
- Not on schedule
- Marketing consent
- Valid
- Recent contact
- None
- Exclusions
- None
AI extracts the condition from text. The source text stays attached to the decision as evidence.
The goal is not “ask AI to guess what a customer feels.” The goal is to extract a usable condition from text while retaining the supporting source evidence. That is a much more specific and defensible action rule than simply targeting “likely churners.”
Who is in the audience
Customers enter and leave as the data changes.
The audience is not a list someone exports. It is recalculated, so membership follows the rule.
- They become overdue
- They reach a value threshold
- They become newly eligible based on engagement
- They submit a qualifying quiz response
- They match a future AI-extracted CRM signal
- They purchase
- They unsubscribe
- They are contacted too recently
- They fail a condition
- They hit an exclusion
A protective rule
Protect customers with unresolved negative feedback.
When a customer reports a negative delivery or product experience, LiftMiner can suppress that customer from generic promotional audiences until the client has reviewed or resolved the issue.
- A recent negative delivery response
- Exists
- A recovery case
- Unresolved
- Safety, quality, legal, privacy, fraud, or chargeback
- Signal present
Any one of these is enough to hold the customer back. A sample policy, not a client rule.
- Excludes the customer from generic upsell, reorder, and subscription campaigns.
- Routes to the appropriate client-owned recovery or escalation workflow.
Why it matters
A marketer should not have to remember every Monday to check which customers are overdue to reorder.