# Agency client onboarding: the content handoff checklist

> Agency client onboarding for white-label content: a handoff checklist and copyable template covering claims, exclusions, sources, reviewers and approvals.

Canonical: https://revnu.com/blog/agency-client-onboarding-the-content-handoff-checklist
Published October 31, 2026 by George Jefferson, Co-founder and CEO, Revnu

# Agency client onboarding: the content handoff checklist

Consider a hypothetical agency handoff. A draft for a security software client says the product "meets SOC 2 requirements," but the client has an audit in progress and no report. The agency did not collect that distinction before drafting.

That mistake did not happen in the draft. It happened at kickoff, when the agency collected logins, scheduled a recurring call and moved on without asking what the writer was allowed to say.

Most agency client onboarding advice covers the kickoff call, the shared Slack channel and access to analytics. All of that matters. None of it tells a writer which claims are safe, who owns the sources, who reviews LinkedIn posts, or who can say yes to a prospect email. This post covers that gap: a content handoff checklist for white-label client blogs, LinkedIn drafts and outreach, plus a template you can copy and fill in before the first draft exists.

## Why the content handoff breaks before the first draft

Revision rounds usually get blamed on writers. Read the actual comments on a rejected draft and you'll see a different pattern. "We don't say that." "That customer isn't public." "Legal needs to see anything that mentions security." "Sarah approves LinkedIn, not me." These are not quality problems. They are missing inputs.

The missing inputs are predictable. Agencies collect access (CMS login, Search Console, LinkedIn page admin) and assets (logo, brand colors, a pitch deck). They rarely collect rules. A pitch deck tells a writer what the client wants to be true. It doesn't tell the writer what the client can defend if a prospect challenges it.

White-label work makes this worse. When the post ships under the client's name, or under the agency's name with the client's product in the headline, the writer is making public statements on someone else's behalf. The client's reviewer sees the draft for the first time at the end, when changing it costs the most.

A longer kickoff call won't fix it. Calls produce opinions, not records. What works is a written handoff that the client signs off on before drafting starts, so every later disagreement can be settled by pointing at a line in a document.

Here's my position: if your agency can't name the forbidden claims for a client, you're not ready to write for them. A first draft written without that list is a guess. The client pays for the guess in revision rounds and you pay for it in margin.

## The handoff checklist at a glance

Seven items. Each one removes a category of revision comment.

1. **Positioning in one sentence.** What the product is and who it is for, in the client's words, approved by the person who can change it.
2. **Allowed claims.** The specific statements the writer can make without asking, each with a source.
3. **Forbidden claims.** Statements the client has retracted, cannot prove or is legally barred from making. Include the ones that used to be true.
4. **Audience and exclusions.** Who the content is for, who it is not for, and which industries, company sizes or competitors are off limits.
5. **Brand references.** Which customers, logos, case studies and quotes can be named, with the exact published URL and the original date.
6. **Source ownership.** Who owns each source of truth (pricing page, docs, changelog, customer data) and who answers questions about it.
7. **Reviewers and approvals by channel.** For each of blog, LinkedIn and outreach: who reviews, who approves, how long they get and what happens if they miss the window.

The order matters. Items one to three settle what can be said. Items four and five settle who it is said to and who can be named. Items six and seven settle who is responsible when something is wrong.

You don't need a long meeting to fill this in. Send it as a form or a shared document, ask the client's marketing lead to complete the first five items and ask their operations or product owner to complete item six. Book thirty minutes to resolve the blanks and disagreements, not to collect the basics.

A useful test: could a contractor who has never spoken to the client write a safe first draft from the document alone? If not, the handoff is incomplete.

## Positioning, product claims and forbidden claims

Start with positioning, because every other decision hangs off it. Ask the client for one sentence: "[Product] is a [category] for [audience] that [specific job]." Then ask them to name two phrases they have retired. Companies change positioning more often than their websites reflect, and old phrases leak back in through sales decks and old blog posts. A writer who finds the retired phrase in a three-year-old post will happily reuse it.

Next, split claims into two lists.

**Allowed claims** are statements the writer can use freely. Each needs a source: a pricing page, a documentation URL, a published case study. "Integrates with Slack" is allowed if the integrations page says so. "Cuts onboarding time in half" is not allowed until someone produces the denominator.

**Forbidden claims** are the list most agencies skip, and it's the more valuable one. Write each as a plain sentence the writer must not assert, imply or hint at. Include four kinds:

- Claims the product cannot support today (a capability that is planned, not shipped).
- Claims that were true and no longer are (old pricing, a retired plan, a discontinued integration).
- Claims that need a certification or audit the client does not hold. SOC 2, HIPAA and ISO 27001 are the usual traps.
- Comparative claims about named competitors that nobody has verified.

Two habits keep this list honest. First, date every entry and record who added it, so you can tell a permanent ban from a temporary one. Second, give the list a single owner on the client side. When three people can edit it, it drifts. When one person owns it, you know whom to ask the next time a writer is unsure.

The rule for the writer is simple: if a claim isn't on the allowed list, ask before using it. Silence is not permission.

## Audience, exclusions and brand references

"Founders and operators" is not an audience. A usable audience line names a role, a company size and a situation: "technical founders of launched B2B software companies with 2 to 20 people and no dedicated growth hire." That sentence tells a writer what to assume the reader already knows, which jargon to skip and which examples will land.

Then write the exclusions. They feel negative, and they save the most time.

- **Industries to avoid.** A client selling to fintech may not want a blog post that uses a healthcare example, because their legal team hasn't cleared it.
- **Segments they do not want.** If the client doesn't serve enterprises, a post that implies they do will attract the wrong leads and annoy the sales team.
- **Topics off limits.** Pricing speculation, roadmap hints, layoffs, funding rumors, competitor attacks.
- **Competitors.** Decide whether any competitor can be named, in what context, and whether a link is ever allowed. The safe default is factual passing mentions only, and never a link.

Brand references get their own line because that's where white-label content most often creates real legal exposure. For every customer, logo, quote or metric the writer might use, record four things: the exact public URL, the original publication date, the denominator behind any number, and the attribution wording. A statement like "customers saw a 40% lift" is a draft-killer unless someone can say 40% of what, over what period, for which customer.

If a customer isn't on the approved list, the writer can't name them, even if the agency knows the relationship exists from a call. Private context from a kickoff meeting is not public evidence.

One extra column helps: "approved for which channel." A customer logo cleared for the website isn't automatically cleared for outreach, and a quote cleared for a blog post may not be cleared for a LinkedIn ad.

## Source ownership, reviewer and approvals by channel

A source of truth with no owner goes stale. Name one person per source: the pricing page, the documentation, the changelog, the integrations list, the customer data. Next to each name, record how fast they usually respond and where to ask (a Slack channel, an email alias). When a writer hits a conflict between two sources, the owner settles it and the answer goes into the allowed or forbidden list so the next writer doesn't repeat the question.

Approvals are where white-label work quietly stalls, because each channel has different risk and different people.

**Blog.** Slow, public and indexed. Name an editor who checks tone and structure and a fact reviewer who checks claims. For technical or regulated clients, add a subject-matter reviewer. Set a review window (two business days is a common starting point) and a default: does a silent reviewer mean approval, or does the post wait? Decide in advance. A post that waits indefinitely costs you a publishing slot.

**LinkedIn.** Fast, personal and tied to a named individual. The reviewer is usually the person whose profile it ships from, because they're the one who answers comments. Agree on tone limits (first person, no hot takes about competitors) and whether posts go out from the founder's account, the company page or both. Social platforms change their rules often, so confirm with the client whether anything is published automatically or only drafted for them to post.

**Outreach.** The highest risk per word, because it goes to individual prospects and can damage a sending domain. Name who approves each new sequence, who approves one-off emails, and who handles replies. Replies should always go to a human decision-maker. Also record any sending limits, blocklists (existing customers, active deals) and opt-out handling.

Write the approval chain as a table, not a paragraph. One row per channel, with reviewer, approver, window and default. If a row is blank, the handoff isn't finished.

## A copyable handoff template (hypothetical example)

Copy this into a shared document and have the client fill it in before drafting. The example below is hypothetical: a made-up client, "Northgate," a workflow tool for IT teams.

**1. Positioning.** Northgate is a ticket automation tool for IT teams at mid-sized companies that routes requests and closes repeat tickets. Retired phrases: "AI-powered helpdesk", "zero-touch IT."

**2. Allowed claims (with source).**
- Integrates with Slack and Microsoft Teams (northgate.example/integrations).
- Pro plan is self-serve (northgate.example/pricing, checked 2 March).

**3. Forbidden claims.**
- Northgate is SOC 2 certified (audit in progress, no report yet).
- Northgate replaces an IT team.
- Any claim about an Enterprise plan price.
- Any statement that a feature "works automatically" where a human still approves the action.

**4. Audience and exclusions.**
- Audience: IT managers at companies of 50 to 500 people.
- Exclude: healthcare examples, government buyers, anything about layoffs or headcount reduction.
- Competitors: factual passing mention only, never linked.

**5. Brand references.**

| Reference | Public URL | Original date | Denominator | Channels cleared |
|---|---|---|---|---|
| Case study: Acme Logistics | northgate.example/customers/acme | 14 Jan | 6 months, one team | Blog, LinkedIn |

Anyone not in this table cannot be named.

**6. Source owners.**
- Pricing and plans: Dana (product ops), replies in #agency-northgate.
- Docs and integrations: Luis (engineering), replies within one business day.

**7. Approvals by channel.**

| Channel | Reviewer | Approver | Window | If no reply |
|---|---|---|---|---|
| Blog | Priya (marketing) | Dana (fact check) | 2 business days | Wait, do not publish |
| LinkedIn | CEO, from own profile | CEO | 1 business day | Skip the post |
| Outreach | Sales lead | Sales lead | Same day | Do not send |

Replies to prospects always go to the sales lead.

**8. Sign-off.** Client contact name, date, and the agreement that changes to this document are made in writing.

## Where an AI drafting workflow fits: a short Revnu example

A handoff document only helps if the drafting process reads it. Whether your drafts come from a junior writer, a freelancer or an AI workflow, the checklist should sit in front of the work, not in a folder nobody opens.

Revnu is one example of how that can work. Per its own description, it produces client blogs, LinkedIn posts and outreach under the agency's name, and the agency approves everything before it goes out. That fits the approval-by-channel structure above: the agency's reviewer, not the tool, decides what ships. Revnu's blog drafts go through review, and the review queue gives the agency a place to see prepared work awaiting approval. On the Pro plan (from $99 a month, self-serve), new outbound sequences ask for approval once and prospect replies always wait for a person. That maps onto the rule that replies go to a human.

Two cautions about any drafting workflow, Revnu included. First, treat the forbidden claims list as an input the drafter must be given, not a hope. A fast draft that repeats a retracted claim is just faster at being wrong. Second, social posts from Revnu are drafts with a share link, so the person whose profile it ships from posts it themselves. Confirm that posting step with your client during the approvals section rather than assuming it.

If you want a broader view of when agent-run work beats a human agency model, [a growth agency alternative when you need the work done](https://revnu.com/solutions/agency-alternative) covers that tradeoff. If you're evaluating tools more generally, the [AI marketing agent evaluation checklist](https://revnu.com/blog/guides/ai-marketing-agent-evaluation-checklist-how-to-test-tools-safely) is a useful companion.

## Conclusion

Agency client onboarding is won or lost in the first document you ask a client to sign, not in the kickoff call. Send the handoff checklist before the first draft, get the forbidden claims and approval table filled in, and refuse to write until the blanks are closed. You'll lose a day at the start and save weeks of revision rounds.

If you produce white-label blogs, LinkedIn drafts and outreach for software clients, try running that checklist through Revnu's agency workflow. It drafts client blogs, LinkedIn posts and outreach under your agency's name and holds everything for your approval before it goes out. Start at revnu.com with Pro at $99 a month, self-serve, with no sales call required, and put your first client's handoff document in front of it before the first draft.

## Frequently asked questions

### What should be in an agency client onboarding checklist for content?

Seven items: one-sentence positioning, allowed claims with sources, forbidden claims, audience and exclusions, approved brand references, named source owners, and reviewers and approvals for each channel. Collect these in writing before drafting. Access to the CMS and analytics still matters, but it does not tell a writer what is safe to say.

### How do you handle forbidden claims when writing for a client?

Ask the client to list statements the writer must not make, including retired positioning, unshipped features and certifications they do not hold. Date each entry and assign one owner. If a claim is not on the allowed list, the writer asks before using it. Revnu's agency workflow keeps the agency approving everything before it goes out.

### Who should approve white-label blog posts, LinkedIn drafts and outreach?

Set it per channel. Blogs usually need an editor plus a fact reviewer. LinkedIn drafts are best approved by the person whose profile publishes them. Outreach needs a named approver for each sequence, and prospect replies should always go to a human. Record the review window and what happens if nobody responds.

### How long should agency client onboarding take before the first draft?

Give the client a form or shared document to complete first, then book a short call, roughly thirty minutes, to resolve blanks and disagreements. The timeline depends on how fast the client's product and legal owners respond, so record each owner's usual response time in the handoff and plan around it.

### Can an AI drafting tool follow a client handoff document?

Only if you give it the document as input and keep a human approving the output. Revnu states that it produces client blogs, LinkedIn posts and outreach under the agency's name, with the agency approving everything before it goes out. Provide the forbidden claims list up front and review drafts against it.