How an AI agent can build a revenue-aware email audience with verified identity matches
This workflow gives an AI agent a defined job, a bounded set of records, and a result a person can review. The agent reads the relevant Kit context and matches it with Stripe, applies the rules in the prompt, and keeps the source behind every recommendation. It returns a proposed handoff rather than taking consequential actions on its own.
Can an AI agent build a revenue-aware email audience with verified identity matches?
Yes. Start with the scope, date range, decision rules, and fields that identify the right records. The agent can collect the evidence, compare states or sources, mark conflicts and missing data, and organize the result around the outcome above. A reviewer then checks the matches and judgment calls before approving messages, record updates, bookings, purchases, publishing, or other write actions. The guide below shows the records, boundaries, prompt, and handoff needed for this specific workflow.
Match customers before segmenting
Kit supplies subscriber status, tags, segments, and email history. Stripe supplies customer, subscription, invoice, and payment context. Used together, they can prepare a revenue-aware audience such as active annual customers who have not received an onboarding sequence or former subscribers who should be excluded from a retention message.
Identity matching comes first. Choose a verified customer ID stored in Kit, an exact normalized email address, or another approved key. The agent should report duplicate emails, multiple Stripe customers, and unmatched records instead of selecting the most convenient match. A valid payment does not override an unsubscribe in Kit.
Example starter prompt
Compare Kit subscribers in [segment or tag] with Stripe customers for [product or subscription rule] as of [date].
Match records using [approved key]. For each match, show the Kit subscription state and tags plus the relevant Stripe customer, subscription, or invoice status. Put duplicates, conflicting emails, and unmatched records in separate sections.
Return a proposed audience and exclusion list. Do not add tags, enroll subscribers, change Stripe records, or send email.Review exclusions before additions
Check unsubscribed, bounced, refunded, disputed, canceled, and past-due cases before approving the audience. The business may have different rules for each state, and the agent should apply only the rules supplied in the task.
The final table should include the match key, Kit record, Stripe record, eligibility rule, decision, and reason. Keep proposed Kit tag changes in a separate approval list so a marketing owner can inspect every write.
Questions this workflow answers
Could an agent identify which paying customers belong in a subscriber segment without emailing people who opted out or matching the wrong accounts?
Yes. Stripe supplies customer, subscription, invoice, payment, refund, dispute, and product context. Kit supplies subscribers, email state, tags, segments, and sequence membership. The agent joins the records with a verified customer identifier or normalized email under rules you define. A similar name or company is not a safe match.
State the commercial condition and messaging eligibility separately. “Has an active annual subscription to product A” describes the Stripe side. “Is subscribed, not bounced, and not already in this sequence” describes the Kit side. Add explicit handling for trials, past-due invoices, canceled-at-period-end subscriptions, refunds, disputes, multiple subscriptions, and shared billing contacts. Unmatched and one-to-many records belong in their own queues.
The output should show the match key, records from both systems, relevant dates and states, eligibility rule, decision, and reason. Payment does not create marketing consent. A current customer who unsubscribed remains excluded, and the agent should preserve suppression even when a campaign goal would benefit from adding them.
A marketing and billing owner review the proposed audience, spot-check matches, and resolve uncertain identities. Tag changes, segment updates, and enrollment remain approval-based. The result is a defensible customer segment built from current commercial status and current communication permission, with every exclusion and unresolved match visible before a message can be sent.
The snapshot time should be recorded because billing and subscription states can change between audience review and campaign activation. A final delta check can remove newly ineligible records without adding anyone whose consent has not been reviewed.