If you schedule social media for a client, you know the thread. It starts with "next week's content for your review" and seven attachments. By Thursday it is eleven replies deep. Somewhere in the middle the client wrote "approved, but can we swap the second image," and now nobody is sure whether Tuesday's post is the old version or the new one. The post goes out wrong, or late, or twice.
I should be upfront about where I sit. Planpost is my product and it is pre-launch, so I have no agency case studies to wave around. What I have is an approval workflow I built specifically to kill that thread, and this post walks through exactly what it does.
Why email is where approvals go to die
Email fails at approvals for a boring structural reason: the thing being approved is not in the email. The client looks at a screenshot of a post, pasted next to a schedule that lives in a spreadsheet, described by prose in the message body. Three sources of truth, all slightly out of date by the time anyone reads them.
Version control makes it worse. "Final_v3_ACTUAL.png" is a joke because it is true. When feedback arrives as a sentence like "make the caption punchier and use the other photo," someone has to translate that back into the actual post, re-export the image, and send everything again. Every round trip is a chance to attach the wrong file.
And then there is the worst sentence in client work: "Approved with one small change." Is that an approval? Do you make the change and publish, or make the change and ask again? Retainer relationships have soured over less.
What the client actually needs
Strip the problem down and the reviewer needs four things:
- The post exactly as it will appear, not a cropped screenshot of a draft.
- The time and platform it is scheduled for.
- A way to object to something specific.
- One unambiguous way to say yes.
What the client does not need is another account. This is the part most tools get wrong. If reviewing a caption requires a signup, a password reset, and a tour of somebody's dashboard, the client retreats to email within a week. I don't blame them. Their job is running their business, not learning my software.
So the constraint I designed around was simple: the client should be able to do their entire job from one link, on a phone, with zero onboarding.
How the approval flow works in Planpost
Here is the whole loop:
You draft the post the way you always would: pick the platforms, write the caption, attach media, set the slot. Then you submit it for approval. Planpost generates a review link with a long random token, and that link is the entire client experience. How you deliver it is up to you. Email works fine as a transport layer once it stops being the workspace.
The client opens the link wherever they are and sees the post rendered the way it will actually look, along with when and where it is going out. No login wall. No trial signup. A reviewer who has never heard of Planpost can approve a post thirty seconds after receiving the link.
Comments attach to the thing being discussed. A problem with the image becomes a comment sitting on the image, not a clause buried in paragraph four of an email. If the caption is fine but Tuesday at 9am is wrong, the comment can point at the schedule instead. When the client is done, they pick one of three states: approved, changes requested, or rejected.
That last part matters more than it looks. Email lets everyone avoid the distinction between "yes," "yes, if," and "no." The workflow forces it, which is uncomfortable for about a week and then quietly fixes the ambiguity that used to cause the real damage.
Auto-publish on approve
There is an optional setting I like a lot: auto-publish on approve. With it on, the moment the client approves, the post flips to scheduled and goes out at its slot through the normal publishing queue. Nobody has to notice the approval, log in, and manually release anything. The yes is the release.
With it off, an approval is just a green light and you publish on your own terms.
My suggestion: start with it off for a new client. The first few rounds of any approval process surface questions about the process itself, and you want a human hand on the wheel while that settles. Turn it on once the rhythm is boring.
Small details that carry the weight
A few deliberate choices in how this works.
A post can only have one pending approval at a time. So "which version is the client looking at" has exactly one answer, always. If you revise after changes were requested, you resubmit and the link reflects the current state of the post.
Every decision is recorded: what was decided, by whom, and when. When someone asks in October why a post went out in September, you have the receipt instead of an inbox archaeology session.
If a post is still waiting on approval 24 hours before its scheduled slot, the reviewer gets a reminder automatically. Chasing sign-off is the least dignified part of running social for someone else. It deserved to be automated.
And links can be revoked. If a campaign is cancelled or your contact leaves the company, you can kill the link instead of hoping nobody clicks it.
The honest caveat
Planpost is pre-launch, so there are no customer quotes in this post and no invented ones either. The workflow above is what the product does today; whether it survives contact with real agency clients is exactly what I am about to find out.
How I test it without clients
The way I exercise this flow is unglamorous. I submit my own posts for approval and review them from my phone, pretending to be a difficult client. I reject things. I request changes on my own captions, which is a strange kind of arguing with yourself, and I recommend it as a product exercise if you ever build software.
That testing convinced me the link is the product. Every step I removed between "client receives thing" and "client approves thing" made the loop faster, and the login screen was the biggest step of all. Nothing else came close.
Where this fits in the week
Approvals sit on top of ordinary scheduling, so if your calendar is chaos, the review link will faithfully display the chaos. Get the base layer right first: how scheduling actually works and building a content calendar in one afternoon cover that ground.
The rhythm I would aim for with a client is weekly. Batch the week's posts, submit them all, send the links in one message on Friday, and collect decisions by Monday. One touchpoint instead of a rolling thread. If a client wants to review daily, the workflow handles that too, but weekly batches are where the time savings live. Planpost comes with a 7-day free trial; plan details are on the pricing page.
The email thread was never really about email. It was about the preview, the schedule, and the decision living in three different places. Put them behind one link and "approved" finally means one thing.