A Client Approval Workflow for Design Agencies

Solo freelancers can sometimes get away with informal approval: one client, one relationship, a shared memory of what was agreed. Agencies can't. With multiple clients, multiple projects, and multiple stakeholders per project, "I think we agreed on this" stops being viable. You need a workflow that produces a record, consistently, across everything in flight.

Here's a repeatable one.

The problem agencies actually have

It's not collecting feedback. Most agencies have a tool for that. It's three things feedback tools don't solve:

1. Multiple approvers per project. A client's marketing lead loves it. Their CEO hasn't seen it. Who actually signed off? On a feedback tool, the answer is "someone left a positive comment," which protects no one.
2. No record across projects. When you're running ten projects, you can't hold in your head which version each client approved and when. You need it logged, per project, retrievable.
3. Scope disputes at scale. A solo freelancer eats one unpaid revision round. An agency running ten projects eats ten. Scope creep that's a nuisance solo becomes a margin problem at agency scale.

The first one is the quiet killer. Most agency approval disputes aren't "nobody approved it." They're "the wrong person approved it." A junior stakeholder said yes, the work shipped, and the decision-maker resurfaces later wanting changes. Without a named approver per deliverable, you have no defence, because the approval you have on record came from someone who didn't have the authority to give it.

The workflow

1. One named approver per deliverable

Before review starts, establish who has approval authority for each deliverable. Not "the client", a specific person. If the CEO needs to approve, the marketing lead's comment doesn't close the round. This prevents the most common agency dispute: approval from someone who turned out not to have authority.

2. Present frozen versions, not rolling drafts

Each round is a specific, named, frozen state, "Homepage, Round 2, for approval", so there's no ambiguity about what's under review. Across ten projects, this naming discipline is what keeps versions from blurring together.

3. Run feedback in context, resolve to zero

Pin comments on the actual work, resolve them one by one, drive each round to zero open items. Standard practice, but the point is that resolving feedback is not the same as closing the round. See feedback vs approval.

4. Close with a formal sign-off from the named approver

When items hit zero, the named approver formally signs off on that frozen version. This is the step that produces the approval receipt, version, approver, date, scope, and it's the step that makes the round defensible.

5. Log every receipt centrally

For an agency, the receipt can't live in one person's inbox. It needs to be retrievable per client, per project, so that any team member can answer "what did this client approve and when?" without archaeology. This is also what protects you when the team member who ran the project has left.

6. New requests open new, scoped rounds

Because each round closed on the record, post-sign-off requests are demonstrably new, which is what lets you scope and bill them rather than absorb them. Across a full project load, this is the difference between protected margin and silent leakage.

What this changes at scale

The per-project version of this workflow is just good practice. The agency version is a financial control. When you run ten concurrent projects, the difference between "we close every round with a logged, named sign-off" and "we mostly remember what was approved" is several unpaid rounds a quarter, plus the occasional larger dispute that costs a relationship.

The workflow doesn't slow you down once it's in place. The naming and sign-off steps take seconds, and they replace the much slower work of reconstructing what happened months later from scattered messages. Discipline up front buys you out of archaeology down the line.

Why this needs tooling at agency scale

A freelancer can run this on signed PDFs. An agency can't. The central log, the per-project retrievability, and the multi-stakeholder sign-offs are exactly the parts manual methods drop under load.

This is what Lyba's agency dashboard is built for: managing sign-offs across multiple clients, with approval receipts recorded per project, so the closing step scales with your project count instead of breaking at it. The feedback loop is the easy part. Every tool has it. The accountability layer across a full client roster is the part that's usually missing, and it's the part that protects agency margin.

FAQ

What's the difference between an agency approval workflow and a freelancer's? Mostly scale and stakeholders. A freelancer often has one approver and can rely on memory. An agency has multiple approvers per project across many projects, so it needs named approvers, frozen versions, and a central, retrievable log of approval receipts rather than informal agreement.

How do I handle multiple stakeholders approving one project? Assign one named approver with authority per deliverable before review starts. Other stakeholders can comment, but only the named approver's sign-off closes the round. This prevents the common dispute where a junior approved and a decision-maker later objects.

Where should approval records live for an agency? Somewhere central and retrievable per client and per project, not in one person's inbox. Any team member should be able to answer "what did this client approve, and when?" instantly, which is why agencies move off manual PDFs onto a shared dashboard.

Does this slow the team down? No. Naming an approver and capturing a sign-off take seconds per round and replace the far slower work of reconstructing decisions later. At scale it protects margin without adding meaningful overhead.

Related: How to get a client to formally sign off on a design · How to handle scope creep · What is a design approval receipt?