Client Support Retainer Termination Offboarding Pack (With AI)

Your client chose not to renew—or you agreed to end a support retainer early. That moment is not a live P1 incident, an SLA make-good, or the commercial renewal conversation you run at term end. It is a controlled wind-down: close open work honestly, hand off what the client owns, revoke access on schedule, and reconcile the last invoice without surprise charges.

This playbook is a retainer termination / offboarding pack for freelancers and small agencies running monthly support retainers. You will use ChatGPT to structure facts, Google Docs for a single checklist the client can sign off on, and Gmail for a short, professional sequence—human sends every message.

If you are still trying to extend the contract, use the renewal pack first. If the retainer is active and healthy, the monthly review cadence still applies until the termination date.

When to run this pack

Run it when:

• The client confirms non-renewal after your renewal memo (or gives written notice per contract).
• You terminate for cause or mutual agreement with a fixed last day of support.
• A new vendor takes over and you need a documented handoff window (typically 5–15 business days, per contract).

Do not use this pack for:

• Active outages (urgent incident pack).
• SLA breaches you caused (service credit pack).
• Selling a retainer after project delivery (proposal pack).

What you will produce

1. Wind-down snapshot — open tickets, hours used vs bank, access list, known risks.
2. One-page Google Doc — termination checklist, handoff index, access revocation dates, billing note.
3. Gmail sequence — notice acknowledgment, handoff confirmation, final closure (three emails max).

Step 1 — Pull a FACT-only snapshot

Export from your helpdesk or spreadsheet:

• Open tickets with last update, owner, and blocker.
• Hours consumed this cycle vs included bank (and any pending top-up approvals).
• Systems where you still have admin access (hosting, DNS, analytics, inboxes, Slack, etc.).

Paste into ChatGPT with this prompt:

You are my retainer offboarding assistant. Separate FACT (quoted from the export) from ASSUMPTION (inference). List open items by severity for wind-down. Do not invent ticket IDs. Output: (1) executive summary, (2) open work table, (3) access inventory, (4) billing flags.

Review every ASSUMPTION line before it goes client-facing.

Step 2 — Draft the termination checklist (Google Doc)

Ask ChatGPT to turn the summary into a client-ready outline:

Turn the wind-down summary into a one-page Retainer Termination Checklist with sections: Last day of support, Open items closure plan, Handoff deliverables (links/files client receives), Access revocation schedule, Final invoice line items, Contacts after offboarding. Use neutral tone. Mark anything needing client input with [CLIENT].

Create a Google Doc from that outline. Keep it scannable: checkboxes, dates, and named owners (client vs you).

Include a handoff index table:

Item	Location / owner after termination	Your action before last day
Runbooks & credentials	Client vault	Export + rotate shared secrets
Open bug #…	Client or new vendor	Close or reassign with written note
DNS / hosting	Client admin	Document current state; remove your user on date X

Step 3 — Close or transfer open work

For each open ticket:

1. Decide: close, hand off (with repro steps), or billable CR (only if contract allows post-notice scope).
2. Log the decision in the Doc—clients dispute “we thought you would finish X” when X was never written down.
3. If a new vendor is involved, BCC is not enough; use explicit written permission to share artifacts.

This step is where retainers quietly fail: silent partial work. The checklist is your proof of professionalism.

Step 4 — Gmail sequence (human sends)

Email 1 — Acknowledge termination & share the Doc

Subject: Support retainer wind-down — checklist and dates

Hi [Name],
Thanks for confirming the retainer ends on [date]. Attached is our one-page wind-down checklist covering open items, handoff files, access removal, and final billing.
Please reply with corrections by [date] so we can align on the last day of included support.
— [You]

Email 2 — Handoff complete / pending client actions

Send when exports are ready or when you need client keys/approvals:

Hi [Name],
Handoff bundle: [links]. Remaining open items: [list with owners]. We will revoke [systems] on [date/time timezone] unless you request a different window in writing.
— [You]

Email 3 — Closure & final invoice

Hi [Name],
Support access has been removed as of [timestamp]. Final invoice [number] covers [period/items]. Thank you for the engagement; here is how to reach us for ad-hoc work if needed: [policy or decline].
— [You]

Draft variants in ChatGPT, then paste into Gmail and edit names/dates.

Step 5 — Access, archives, and final billing

On the agreed revocation date:

• Remove your users from client systems; keep an internal log (no passwords in email).
• Archive ticket exports and the signed checklist PDF in your records.
• Reconcile hours: unused bank policy should match what you already documented in unused hours or the signed SOW—do not improvise at exit.

If the client disputes scope, your Doc + email thread is the audit trail—not a Slack reaction emoji.

Step 6 — Optional retrospective (internal)

If the relationship ended abruptly, run a blameless internal note (not necessarily the full post-incident RCA): what signal you missed, whether pricing or SLA clarity failed, one process tweak before the next retainer kickoff.

Copy-paste ChatGPT starter

Context: Monthly support retainer ending [date]. Contract notice period [N days]. Open tickets: [paste]. Hours: [used/included]. Access: [list].
Task: Produce wind-down checklist content for a Google Doc and three short Gmail drafts. Label FACT vs ASSUMPTION. Flag legal/billing items I must verify.

Related guides

Browse more client ops playbooks in Guides. For starting retainers cleanly, see kickoff.

---

You approve every client email and every access change. AI drafts; you remain accountable.