The Approval Budget: What We Changed After Asking One Founder for 130 Sign-offs

In August 2026 we planned the second month of content for MH Legal Hub, an immigration law firm in Buenos Aires whose founder answers personally for every sentence the firm publishes. Four guides, films, posts and clips, all of it checked against the law, with each legal claim written out and flagged for review.

Nobody had added the flags together. When we did, the plan needed roughly 130 legal approvals from one person, a lawyer with clients waiting. A plan that cannot ship until 130 approvals land is a queue with a calendar drawn on it.

The instinct in that situation is to chase. Chasing is the wrong response, because the founder was doing the right job. Clients come first, and content approvals come after them, where they belong.

Why nobody sees it coming

Each flag on the list was responsible work: a real legal claim, honestly surfaced, correctly worded. The defect only exists in the total, and nothing in the drafting process ever sums it.

Careful writing makes it worse. The better a team is at noticing what needs a lawyer’s eye, the longer the list gets. A careless team would have sent fewer questions and published more errors. A careful team without a budget sends a list that nobody can finish.

Production hours stretch. A team can work late, add a person, reuse a layout. The founder’s attention does not stretch, and it was the one input in the plan that nobody had counted.

What we changed

We now treat a client’s decisions as a budget, set before the deliverables are chosen.

Count the asks before the assets. If the number passes about a dozen, cut the plan. Long reference documents cost the most, because they are full of stated claims. A page built from the client’s already approved words costs almost nothing.

Rank by what each answer unblocks. An unsorted list of 67 flags says everything matters equally, and it reads as homework. Put first the answers that stop a whole piece. Some of the rest can wait, and some turn out not to need asking.

Reuse what the client has already said. The founder’s own posts, corrections and voice notes need confirmation, and confirmation is much cheaper than authorship.

Bring a recommendation. “Is this correct, yes or no, and here is our reading” takes a lawyer seconds. “What should we say about this?” takes an afternoon.

Shrink the ask. A request to block ninety minutes went unanswered. “We are filming on Tuesday anyway, say yes to the camera” landed where that request had not. The size of the decision was what changed.

A client’s attention is the only part of a content plan that does not stretch. Count it first.

Give the approvals a room

The second half of the fix was where the decisions happen. Approvals that live in email threads and chat messages get lost, and they get answered in fragments.

The firm now has one review room. Each piece carries three answers: Approve, Send feedback, Decline. A note can be pinned exactly where it applies, on a slide, a moment of a video, a spot in an image or the words of a caption, so the founder never has to describe a location in a sentence.

Three rules keep the room honest:

• Not approved means not scheduled. The calendar does not move anything forward on the assumption that silence is a yes.
• Every returned piece says what changed. A short “What changed since your comments” line means the founder reads the difference and does not have to read the whole piece again.
• Feedback applied never turns into approved. When we finish a correction, the piece goes back and waits. Only the founder can move it to approved, because the founder is the one who signs.

Long documents follow the same logic. They open on three key pages, and everything else sits behind a pointer of the form “Details: section 4, page 12”. A founder who reads only the three pages can still decide.