Agency operations playbook

Proxy Access Control SOP | NobleProxy

A copyable SOP for agency teams: every endpoint has exactly one accountable operator, and access matches the roster - with a named owner, a trigger, a checklist, and measurable signals.

Quick answer

Every endpoint has exactly one accountable operator, and access matches the roster. Owner: Team lead. Trigger: On every roster change; full matrix review monthly. Work the checklist top to bottom and record every change with a date.

Named owner and trigger
Copyable checklist
Escalation path
Measurable signals

When this is the right fit

  • Named owner and trigger
  • Copyable checklist
  • Escalation path

Who this page is for

A copyable SOP for agency teams: every endpoint has exactly one accountable operator, and access matches the roster - with a named owner, a trigger, a checklist, and measurable signals.

Every endpoint has exactly one accountable operator, and access matches the roster. Owner: Team lead. Trigger: On every roster change; full matrix review monthly. Work the checklist top to bottom and record every change with a date.

The checklist

  • 1Keep an endpoint-to-operator matrix with one accountable operator per endpoint
  • 2Grant access on roster join; revoke within one business day of a departure
  • 3Require a dated reason for any shared-use exception
  • 4Reconcile the matrix against actual usage logs monthly
  • 5Archive revoked pairs instead of deleting them

Trigger: On every roster change; full matrix review monthly. Cadence: as triggered, reviewed monthly.

Playbook facts

ObjectiveEvery endpoint has exactly one accountable operator, and access matches the roster
OwnerTeam lead
TriggerOn every roster change; full matrix review monthly
CadenceOn trigger; monthly review
Records keptDated log entry per change

Measurable operational signals

  • 1Every active endpoint maps to one operator
  • 2Departure revocations completed within one business day
  • 3Monthly matrix reconciliation has a dated record

Sources

Facts last reviewed 2026-09-01.

Escalation path

If a checklist step fails twice, the operator escalates to the team lead the same day. The team lead decides between a rollback and a scheduled fix, and records the decision in the runbook. Unresolved after 48 hours: escalate to the vendor's support with the incident notes attached.

195+ countries
99.9% uptime guarantee
Unlimited bandwidth
4.9/5 from 127 ratings

Frequently asked questions

Can two operators share one proxy endpoint?

Only as a dated exception with one accountable owner. Shared accountability is how unused endpoints stay unnoticed and billed.

How fast should access be revoked when someone leaves?

Within one business day, and the matrix should record the revocation date next to the endpoint.

Choose one stable proxy for every active profile

Buy self-serve now, or talk with us about a larger agency inventory.