Social Media Post Approval Workflow

An approval workflow is the step between a post being written and that post reaching somebody else’s audience. It sounds procedural until you have skipped it once. The version most teams actually run is a screenshot pasted into a chat thread, a “looks good” three days later, and a scheduled post that goes out on Tuesday whether or not the reply ever arrived.

The failure is rarely the writing. It is the join between the review and the queue. Approval happens in one system, publishing happens in another, and nothing connects them. A post approved in WhatsApp still has to be found in the scheduler and released by hand, and a post that was never approved publishes anyway, because the scheduler was never told to wait. When somebody says a client post went live unapproved, this is almost always what happened. Nobody ignored the process. The process had no way to stop the post.

A real approval workflow closes that join. Review status lives on the post, publishing reads that status, and an unreviewed post simply does not go out.

Approvals are usually described as an agency-to-client mechanism, and that is one use. The other is the person on your own team whose work should be read before it goes out. The setup for that is a single toggle, not a second product.

Open Settings, then Workspace, then the Users tab, and click the shield icon next to the person. The dialog groups every permission into five categories: Posts (create, edit, delete, schedule, approve), Comments (internal team comments), Accounts (connect accounts, generate connect links), Filter Tags (create, update, delete) and Workspace (settings, managing members, API keys, deleting the workspace). Toggle Approve posts off for that one person and save. An Override badge then marks the permission as customised away from the role default, so you can see at a glance who has been adjusted.

What changes for somebody without that permission:

• Everything they create is held at Pending approval, whatever they selected in the composer.
• That covers posts they write directly, posts they duplicate from an existing one, and every row of any bulk CSV import they run.
• Publish now disappears from their composer, because publishing immediately would step around the review.
• Nothing else changes. They still create, edit, schedule, delete, comment and tag exactly as before.

Nobody has to watch the calendar for work that has appeared. Submitting posts for approval raises an Approval requested notification, and it goes only to the people in that workspace who actually hold Approve posts, never to the person who submitted. A batch raises one notification rather than one per post, so scheduling a month does not bury a client under fifty pings. Changing an approval status notifies in the other direction as well.

This is the shape to reach for when a new hire is producing good work at a pace nobody has time to pre-read. Leave them as an Editor with full run of the calendar, take away one permission, and the review becomes structural rather than a habit somebody has to maintain. The full walkthrough lives in roles and permissions in the help centre.

Two different access problems get muddled together here. One is letting a client review your work, the other is getting their accounts connected. Neither should involve a password changing hands.

For reviewing, invite them. From the Users tab, Add User takes an email, a first and last name, and a role, where the choices are Editor or Client. They receive an emailed invite link. If they already have a PostFast account it is a single Accept; if they do not, they set a password in the same step. Then they are signed out once and asked to log in fresh so the new permissions take effect cleanly. If the email does not arrive, the three-dot menu on their row generates a shareable link you can send through any channel.

For connecting accounts, send a connect link instead. Generate one from Accounts, choose how long it stays valid (1 to 30 days), optionally scope it to specific platforms, and the client authorises their own accounts on each platform’s own login screen. The scope is enforced server-side rather than in the interface, so a link restricted to Instagram cannot be talked into connecting anything else. The client does not need a PostFast account to use one, and PostFast can email the link on your behalf. The same thing exists on the REST API if you would rather generate links from your own onboarding flow.

The honest caveat, since it affects what a roster costs: a client you invite occupies one of that workspace’s user slots. Members are counted regardless of role, so a Pro workspace with four of your staff in it has room for eleven more people, client reviewers included. That is a real limit and the section on cost below prices it out.

A review gate is only worth having if it cannot be walked around, and the ways around one are the bulk paths: somebody uploads a spreadsheet or wires up an integration, and three hundred posts appear in the calendar having met nobody.

Bulk CSV import takes up to 200 rows and a 10 MB file per upload, covering content, schedule time, target accounts and publicly hosted media URLs. When the person running the import does not hold the Approve posts permission, every row lands at Pending approval. Importing a quarter of content is not a way to bypass a reviewer.

The REST API exposes the same field the composer does. Create a post with approvalStatus set to PENDING_APPROVAL and it waits for a human, exactly as a post typed into the app would. The API is on every plan, including the trial, so a pipeline that drafts content automatically and hands it to a person to release is available at the entry price rather than at an enterprise tier.

AI assistants are the same story with a friendlier surface. Connect ChatGPT or Claude to the MCP server and ask for a client’s week in their voice, and the drafts can land as Pending approval in that client’s workspace. The review step stays where it was. What changes is who does the typing, not who does the signing off, which is the correct division of labour for anything going out under somebody else’s name.

Most teams reading this already have an approval process, held together by a chat thread and somebody’s memory. It mostly works, which is why it survives, so it is worth being precise about what it costs rather than dismissing it.

The review is detached from the queue. Approval in a thread has no effect on the scheduler. Somebody has to carry the decision across by hand, and the post publishes on time whether or not they did. Every unapproved post in that model is live by default and held only by a human remembering.

The record is scattered. Who approved what, and when, is spread over an inbox, a phone and a spreadsheet tab. When a client asks why something went out, reconstructing the answer takes longer than the post did.

Context goes missing. A screenshot is not the post. It does not carry the scheduled time, the target accounts, the first comment or the character count, so clients approve something subtly different from what actually publishes.

It does not scale past a few clients. One thread per client times a roster is a full-time job of chasing, and the chasing is unbillable.

Moving the review onto the post fixes the first point outright, and that is the one that causes the incidents. Nothing here stops you nudging a client in a chat. The difference is that the nudge is no longer what holds the post back.

Approvals are not priced separately. Every plan carries the roles, the per-user permission overrides and the publishing gate, starting at Starter, EUR 10/mo on annual billing (EUR 12/mo monthly). The trial has them too, so you can test a real sign-off loop with a real client before paying anything.

What the plan actually buys is capacity. Seats and workspaces, per plan, on annual billing:

• Starter, EUR 10/mo annual (EUR 12/mo monthly): 1 workspace, 1 user. Enough to schedule, not enough to run a review loop with somebody else.
• Creator, EUR 24/mo annual (EUR 29/mo monthly): 4 workspaces, 5 users per workspace.
• Growth, EUR 40/mo annual (EUR 49/mo monthly): 10 workspaces, 8 users per workspace.
• Pro, EUR 82.50/mo annual (EUR 99/mo monthly): 30 workspaces, 15 users per workspace, 120 social accounts.
• Enterprise, EUR 199/mo annual (EUR 239/mo monthly): 110 workspaces, 30 users per workspace.

The claim to hold onto is no per-seat fee, not unlimited people. Adding a fifth reviewer to a Growth workspace costs nothing extra; adding a ninth is not possible without moving to Pro. And because an invited client counts as a member like anybody else, a workspace with three staff and two client-side approvers has used five of its slots.

Worked through: an agency of five people running ten clients on Pro, one workspace each, pays EUR 82.50/mo on annual billing and has ten more slots in every one of those workspaces for the client’s own reviewers. The full table is on the pricing page, and the agency arithmetic against the per-seat suites is worked out on the agencies page.

A comparison where the author wins every row is worth nothing. Four places the tools above are genuinely better, and one thing PostFast does not do at all.

Sendible does not cap users, on any tier. Unlimited users from Core at $30/mo annual upwards, and client dashboards where the client previews and approves are on every tier including that one. If your constraint is the sheer number of people who need to look at something, that is a better fit than a 15-slot workspace. Assignment and approvals as a routed step start at Plus, $84/mo annual.

SocialPilot gives client seats away. From Premium at $85/mo annual, client seats are unlimited and free, and they sit outside the user count rather than eating into it. That is exactly the thing PostFast does not do: our client reviewers consume a workspace slot. If your model is many client-side approvers per account, SocialPilot prices it better.

Buffer is cheaper at small scale. Team is $10 per channel per month billed yearly with unlimited team members, so three channels and six reviewers is $30/mo. Per-channel pricing inverts as the roster grows, and a thirty-channel agency is paying $300/mo, but for a small brand team the entry point is lower than ours.

Hootsuite Advanced bundles more than an approval gate. At $399 per user per month it is the most expensive way on this page to get a sign-off step, and post approval workflows are marked as not included on both cheaper tiers. What that price does buy alongside it is message auto-routing, customer-care coordination and team performance reporting, none of which PostFast has.

And a limitation of ours worth stating plainly. PostFast’s approval is a single gate rather than a routed multi-stage chain. There is no sequence that requires an editor to sign off before the post is passed to the client as a second mandatory stage. Agencies run those two rounds by convention and it works, but the order is not enforced by the product. If a hard multi-step chain with named stages is a procurement requirement, this is the wrong tool and it is cheaper to know that now.