Minimum data to note (source of truth and matching keys)
- Source of truth for each record type (e.g. CRM for contact details, accounting for invoices).
- Exact field names and formats to sync (use the system field name where possible: email, customer_id, invoice_number, date_of_birth). Note which fields are mandatory.
- Matching key(s) to join records (email, external_id, UID). Say explicitly whether keys are unique and how duplicates should be handled.
- Who owns the data day‑to‑day and who can edit it (name and role), plus the location where updates must be made.
Operational and testing details to capture
Write down expected volumes and cadence: average and peak records per hour, whether sync is real‑time or batched, and any API rate limits you know about. This helps pick the right integration method and test size.
Describe error handling and testing: what counts as a successful sync, retry rules (how many attempts and intervals), where failed items go for manual review, and what a rollback looks like. Include an easy test plan with a small set of test records and acceptance criteria.
GDPR, access, handover and next practical steps
Record lawful basis for processing, retention rules, whether any fields should be pseudonymised or excluded, and who is the data controller and processor for the flow. Note requirements for subject‑access requests and how you will delete or correct synced data.
Finally, capture the practical handover: admin usernames for staging, a contact for each system, sign‑off criteria, and a short list of follow‑ups after go‑live (monitoring checks and first‑week sample audits). If you want a simple filled template for common pairings (CRM→accounting, booking→calendar) to hand to a developer or Zapier consultant, Optira can help prepare it.