A social media approval workflow for freelancers with one client

Agency approval tooling is built for teams of twelve. Here is a workflow that fits one person and one client without adding seats or software.

If you post on behalf of someone else, they want to see it first. Every tool that solves this properly is built for agencies — approval hierarchies, roles, seats per reviewer — and priced accordingly.

For one person and one client, that is enormous overkill, and the usual fallback is a spreadsheet and an email thread. Here is a middle path.

Why the email thread fails

• No version of record. Which draft did they approve — the one in the email, or the one after their reply?
• No preview. Copy in an email does not show what the post will actually look like, so you get approval on something and ship something else.
• Approval by silence. No reply becomes tacit yes, until the one time it does not.
• Re-keying. You paste approved copy into the scheduler by hand, which is where typos are introduced after review.

What good looks like at this size

Three properties. Everything else is agency machinery you do not need:

• They see the real post, rendered as it will appear, not copy in a document.
• Approval is a recorded action, not an email you have to keep.
• Approved content publishes without re-keying, so what shipped is what they saw.

The seat problem

The thing that makes this expensive is that most tools model a reviewer as a user. Your client needs an account, an invitation, a password, and a seat on your plan — for something they do twice a month.

In practice clients do not complete that setup. They reply to the email instead, and you are back where you started.

"Any approval step your client has to create an account for is an approval step that will not happen."

A review link they open in a browser — no account, no seat — is the difference between a workflow that runs and one that exists on paper. It is why PostDaily does approvals that way, which is disclosure rather than a pitch; the principle holds whatever you use.

A workflow that fits

• Batch a fortnight of posts in one sitting rather than sending them one at a time.
• Send one link covering all of them. One review session beats fifteen notifications.
• Set a deadline in the message — approve by Thursday or it goes as drafted. Removes the silence problem without nagging.
• Keep rejections specific. "Change the second line" beats "not quite right", and it is worth asking for that explicitly at the start of a relationship.

Agreeing the rules before the first batch

Most approval friction is not tooling, it is an unstated expectation discovered late. Five minutes at the start of the relationship prevents most of it:

• Who has final say. One person. If two people review, they will disagree in front of you and you will be the one resolving it.
• What happens on silence. State the default explicitly — approved, or held. Either is fine; ambiguity is not.
• How many revision rounds are included. Without a number, some clients will iterate indefinitely because nothing stops them.
• What does not need approval. Replies, usually. If every reply needs sign-off you do not have a social account, you have a press office.

When approval is the wrong tool

Approval works for planned content. It does not work for anything time-sensitive, and trying to force it there is how accounts go silent during the exact moments they should be talking.

The usual resolution is a standing agreement: scheduled content goes through review, and anything reactive is either pre-authorised within agreed boundaries or simply does not happen. Deciding that in advance is much easier than deciding it at 4pm on a Friday.

Keeping the record

The value of a recorded approval is almost entirely in the one conversation a year where someone asks why a post went out. A timestamped record of who approved what, attached to the post itself, ends that conversation in ten seconds.

Two rounds a month, fifteen minutes each. That is achievable, and it is roughly the ceiling of what a one-client relationship will bear before the process costs more than it protects. The same rhythm works for an account run on behalf of a project, where the reviewer is a colleague rather than a client.