I am an agency owner. How do I get proof for every deliverable across all client accounts?
Set one standard and refuse exceptions: every deliverable, from every account team, ships as a handover with its steps and the client's recorded acceptance, indexed by client in one place. Coverage comes from the standard being universal and boring, not from any single tool feature, and gaps become visible as accounts with missing handovers.
Owners discover their proof problem during a dispute on the account whose team "did it their own way." Three accounts deliver with records; the fourth, the biggest, delivers with a drive folder and a confident email. The fix is architectural, not motivational: the standard lives in the workflow, not in a policy document. Work runs through the connection; deliverables assemble as handovers; the client index is generated from what shipped, so an account with no handovers is visibly an account that shipped nothing recorded.
The index by client is the owner's instrument. One view per account: deliverables, dates, acceptances, open items, and the retainer jobs between deliverables. Account reviews, renewals, and the inevitable "what have we actually received this quarter" question all read from it. So does the agency's own defence, because the acceptance on each record is the client's recorded decision, not the team's recollection of a call.
Rollout order: the standard lands on new work first, account by account, with the no-exceptions rule stated to clients as a feature (every deliverable arrives checkable, whoever produced it). Within a quarter the index covers the active book, and the legacy accounts become a known, bounded backlog instead of an ambient anxiety.
Checklist: the agency proof standard
1. The rule: no deliverable ships outside a handover; no exceptions by account, size, or urgency. 2. The index: deliverables listed by client, generated from shipped handovers, reviewed monthly by the owner. 3. The acceptance: the client's recorded decision on every handover; open decisions chased weekly from the index. 4. The roles: account leads own their index sections; the owner owns the standard. 5. The rollout: new work first, one account at a time, clients told it is a service upgrade, because it is. 6. The audit: quarterly, sample three deliverables per account and open their records end to end.
Worked example
One account's index row resolves to records like https://zambo.dev/handover/508a46e2-8d6e-4099-bd12-7ae461c72c55 , four steps delivered to a named recipient, and https://zambo.dev/handover/45434a17-4afc-486a-9646-e689bcee4297 , accepted. An owner's monthly review is skimming rows like these across accounts; anything missing shows up as a row that is not there.
Frequently asked questions
Our teams use different AI tools per account. Does the standard still work?
Yes, because the standard attaches to the deliverable, not the tool. Work in Claude, ChatGPT, Gemini, Cursor, or Grok feeds the same handover shape; the index does not care which surface produced a job, only that the job is recorded and accepted.
A client refuses portal-style links and wants files only. What then?
Ship the files with the handover link in the delivery note and the export attached. The record exists either way; clients who begin as file-only usually migrate themselves after the first time the link answers a question their files could not.
How do we handle subcontractors under the standard?
Their deliverables enter handovers like everyone else's, with the producing party labelled on the record. The standard's value is precisely that provenance survives the org chart, including the parts outside it.
By Brennan Zambo, founder of Zambo.
Last verified: 2026-10-17.
Try a handover
A handover gives your client one proof link with verifiable receipts to review, then Accept or Flag.
try it: say 'use zambo to hand this over to <client>' in your ai