An approval workflow is the path a post takes from idea to published. Most teams never draw that path, so it lives in someone’s head, in three Slack threads and a spreadsheet that nobody updates. The result is the same every week: a post sits in limbo because nobody knows whose turn it is, and the Friday batch goes out late or half-reviewed.
This guide is about the mechanics of approval itself: the stages a post passes through, the states it can hold, who acts at each gate, and the service levels that keep it from stalling. It is deliberately tool-agnostic. You can run this on a Trello board or inside an AI social system like Utin, but the structure has to exist either way.
Stages versus states
These are two different things and confusing them is why approval feels chaotic.
A stage is a step in the pipeline: Draft, Review, Revision, Approved, Scheduled. A state is the status a post holds at any moment: needs work, waiting on you, blocked, ready. A post in the Review stage can be in the state “waiting on legal” or “changes requested.” When everyone can see both at a glance, the question “what is happening with the launch post?” answers itself.
A workable set of post states:
- Drafting — being written, not ready for eyes
- In review — assigned to a named reviewer with a deadline
- Changes requested — sent back with specific notes
- Approved — locked, no further edits without re-review
- Scheduled — slotted into the calendar
- Published — live, now owned by reporting
The rule that makes this work: an approved post is frozen. If someone edits the caption after approval, it flips back to “in review.” Silent edits after sign-off are the single most common way unapproved copy reaches a live feed.
One-step or multi-step?
Not every post needs the same gauntlet. Match the path to the risk.
| Post type | Approval path | Typical turnaround |
|---|---|---|
| Evergreen tip, curated link | Author self-publishes against a checklist | Same day |
| Standard promo, event reminder | Single reviewer (editor or manager) | 1 business day |
| Campaign launch, paid creative | Editor, then brand or marketing lead | 2 business days |
| Claims, pricing, regulated topic | Editor, then subject expert or legal | 3+ business days |
If you push everything through the heaviest path, reviewers drown and start rubber-stamping, which defeats the point. Tiering approvals by risk is the difference between a workflow people respect and one they route around. A separate content review checklist is what makes the lightest tier safe to self-publish.
Set SLAs or the queue dies
The quiet killer of approval workflows is unbounded waiting. A post “in review” with no deadline can sit for a week. Attach a service level to each gate:
- Reviewers respond within one business day or the post escalates to a backup.
- Authors turn around “changes requested” within one business day too. Approval is a two-way SLA.
- Anything still unactioned after the SLA shows up on a daily “stuck posts” view that someone owns.
Two numbers tell you if the workflow is healthy: median time-in-review (how long a post waits per gate) and first-pass approval rate (share of posts approved without revision). If time-in-review creeps past two days, you have a bottleneck reviewer. If first-pass approval sits below 60%, your briefs are too thin and authors are guessing at what reviewers want.
Make feedback land once
Most revision loops are not caused by bad writing. They are caused by feedback arriving in pieces: legal flags a claim on Monday, the brand lead hates the hook on Wednesday, and the author rewrites twice. Collect all notes on a post before it goes back, and require feedback to be specific and actionable. “Make it punchier” is not a revision instruction. “Cut the first sentence, lead with the 40% stat” is.
A practical convention: reviewers leave comments anchored to the exact line, mark the post “changes requested” only once per round, and the author addresses every comment or replies why not. This single discipline routinely cuts a three-round cycle to one.
Where the source belongs
Approval is faster when a reviewer can see where a claim came from without asking. If a draft was built from a pricing page, a case study or an FAQ, keep that link attached to the post. The reviewer checks the source instead of pinging the author, and the comment thread stays short. This is the practical reason website-to-social media strategy and approval reinforce each other: source-led drafts are simply easier to approve.
A reviewer’s pre-approve scan
Before clicking approve, a reviewer should be able to confirm in under a minute:
- One clear message, right audience.
- Claims are supported by an attached or linked source.
- Tone matches the brand and the channel.
- The CTA fits where the reader is in the journey.
- Links, tags and mentions resolve correctly.
- Nothing here needs a higher-tier sign-off it skipped.
If two or more fail, send it back rather than approving with caveats. Caveats get lost; states do not.
How Utin fits
Utin is being built so this whole pipeline lives in one place. It scans your website, drafts channel-specific posts with the source attached, then carries each post through visible stages and states with reviewer SLAs and a “what is stuck” view. The goal is that approval stops being a chase and becomes a queue that clears itself. If you want to follow along, you can register interest in an early pilot.
Approval is one layer of a larger operating system. For the people side of who does what, see social media team workflow . For org-wide rules and access, read social media governance . When AI drafts the posts, pair this with AI posts with human review .