# Control-Tower Readiness Checklist

Confirm the organization has agreed on the business decisions, data definitions, owners, and
recovery workflows the interface must support — before selecting or building a control tower.
Published by Operational Visibility (operationalvisibility.blog).

## Priority decisions and exceptions

- [ ] The top 5–10 operating decisions the tower must support are written down.
- [ ] The exceptions that create the most cost, delay, risk, or customer impact are identified.

## Canonical objects and identifiers

- [ ] Core business objects (order, shipment, receipt, return, invoice) have canonical
      definitions.
- [ ] Correlation identifiers are mapped across systems (marketplace ID, commerce ID, EDI control
      number, warehouse ID, tracking, invoice).

## Source and metric lineage

- [ ] Every launch metric has a definition, source system, transformation owner, and drill-through
      path.
- [ ] Disputed metrics have a named decision owner and resolution process.

## Data freshness and quality

- [ ] Each feed has a freshness target and a visible last-successful-update indicator.
- [ ] Data-quality checks exist for the fields that drive prioritization.

## Ownership and escalation

- [ ] Every exception type has a named owner or governed queue.
- [ ] Escalation paths and response deadlines are defined before incidents occur.

## Business service levels

- [ ] Service levels are defined on completed business outcomes (acknowledgement, shipment
      confirmation, invoice acceptance), not only API uptime.

## Retry, idempotency, and rollback

- [ ] Replays are idempotent; duplicate detection uses stable identifiers.
- [ ] The system verifies business state before retrying.

## Security and access

- [ ] Access is role-based; actions requiring approval are gated by policy.

## Executive and operator views

- [ ] The executive view shows trustworthy summaries with lineage one click away.
- [ ] The operator view shows evidence, expected next event, owner, and deadline per exception.

## Outcome measurement

- [ ] Value is measured through avoided impact: chargebacks avoided, orders recovered, labor hours
      reduced, time-to-detection improved.

Source article: https://operationalvisibility.blog/articles/control-towers-need-exception-economics