Trace every lead entry point
Diagram instant-form submission, API retrieval, automation middleware and CRM creation. Record each system's event timestamp, platform lead identifier, contact fields and failure handling. Use a staging test rather than a live customer record.
Practical checkpoint: assign a named owner, preserve a dated baseline and review a sample of at least ten relevant records where volume permits. Document the expected result, the observed result and any exception before approving the next change. For meta ads, decisions should follow business outcomes and evidence rather than a single platform metric.
Define the identity and deduplication policy
Choose a stable source event ID as the primary deduplication key. Do not merge unrelated people merely because they share a household email or telephone. Define how long a repeated inquiry remains the same lead and when a genuinely new sales cycle begins.
Practical checkpoint: assign a named owner, preserve a dated baseline and review a sample of at least ten relevant records where volume permits. Document the expected result, the observed result and any exception before approving the next change. For meta ads, decisions should follow business outcomes and evidence rather than a single platform metric.
Audit webhook retries and idempotency
Test the same payload twice and simulate delayed delivery. The CRM should create one inquiry and log subsequent deliveries as retries. Store credentials securely and limit payload retention; hashing a contact field alone does not authorize unrestricted reuse.
Practical checkpoint: assign a named owner, preserve a dated baseline and review a sample of at least ten relevant records where volume permits. Document the expected result, the observed result and any exception before approving the next change. For meta ads, decisions should follow business outcomes and evidence rather than a single platform metric.
Separate new inquiries from reopened leads
Distinguish duplicate submissions, existing customers, reopened opportunities and newly qualified prospects. Record all statuses without making each one a primary conversion. Sales should be able to override a mistaken merge with an audit trail.
Practical checkpoint: assign a named owner, preserve a dated baseline and review a sample of at least ten relevant records where volume permits. Document the expected result, the observed result and any exception before approving the next change. For meta ads, decisions should follow business outcomes and evidence rather than a single platform metric.
Reconcile Meta and CRM totals
Compare Meta leads, successfully retrieved leads, CRM inserts, unique people and qualified outcomes by submission date. Expect attribution and timing differences. Investigate exceptions before asserting that one dashboard is wrong.
Practical checkpoint: assign a named owner, preserve a dated baseline and review a sample of at least ten relevant records where volume permits. Document the expected result, the observed result and any exception before approving the next change. For meta ads, decisions should follow business outcomes and evidence rather than a single platform metric.
Operationalize quality feedback
Set an owner for failed webhooks and missing leads, alerts for sudden drop-offs, and a weekly sample audit. Only send downstream feedback through supported, consent-compliant integrations. Review lead quality, not just platform CPL.
Practical checkpoint: assign a named owner, preserve a dated baseline and review a sample of at least ten relevant records where volume permits. Document the expected result, the observed result and any exception before approving the next change. For meta ads, decisions should follow business outcomes and evidence rather than a single platform metric.