How to stop chasing clients for approvals

You sent the homepage mockup on Tuesday. It is Friday afternoon, the client has not replied, your developer is sitting on a build she cannot start, and you are typing the fourth version of a message that begins "just circling back." The project is not blocked by work. It is blocked by a decision nobody has made, and the cost of that decision lands on you, not on the client.

Why client approvals stall

Approvals stall because the request asks the client to do too much thinking in a format built for skimming. It is almost never that they do not care about the project.

Four things cause most of the delay:

• Too many decisions in one message. An email covering copy on three pages, two logo directions, and whether the phone number belongs in the header is not one approval. It is six. The client opens it between meetings, senses it needs twenty focused minutes, and mentally files it under later. Later has no calendar slot.
• The ask is not actually an ask. "Let us know your thoughts" gives the client no idea what a finished reply looks like. Ambiguity is expensive to resolve, so it gets deferred.
• No single place to answer. The mockup is a link, the copy is a PDF, last week's feedback was a text message. The client has to reassemble the context before responding.
• You asked someone who cannot approve it. The office manager can gather feedback but cannot sign off on the brand. She forwards it to the owner, who never received it as a request, only as an attachment.

None of these are personality problems. They are structural, which means you can fix them from your side without changing the client.

How to structure an approval so it takes one click

An approval takes one click when the client faces exactly one decision, has a stated default, and can record the answer without composing a message.

Six rules do most of the work:

1. One decision per request. If answering would take you more than ninety seconds in the client's chair, split it.
2. Show the thing in the message. A preview image beats an attachment. An attachment beats a link that requires a login the client has never used.
3. Write the ask as yes/no or A/B. Never open ended. "Approve this layout" or "A or B" is answerable from a phone in a parking lot.
4. State what happens next either way. Approval triggers something specific. So does a change request. Say both.
5. Give a date and a default. The default is what you will do if you hear nothing, and it should be genuinely reversible at that stage.
6. Name one approver. Put their name in the first line. If you are not sure who it is, that is the real question to ask first.

The approval request template

Subject: Approve: homepage layout (one decision, 2 minutes)

Hi Dana, one decision for you.

What: Homepage layout, version 2. Preview below.

The question: Does the hero section lead with the appointment booking button, or with the new patient offer? Everything else on the page is locked.

Option A: Booking button first. Best if most of your calls are existing patients rescheduling.
Option B: New patient offer first. Best if you want more first time visits.

Reply with A or B. That is the whole thing.

If approved: We build it Thursday and you see it live on staging Monday.
If I do not hear back by Thursday: We build option A, which is reversible for free until we start the mobile pass on the 19th.

Notice what is missing. No status update, no summary of last week's call, no second question hiding in the last paragraph. Those belong in a separate message, because anything you add to an approval request lowers the odds of getting the approval.

The psychology of the polite nudge

A follow-up works when it removes the client's embarrassment about the delay instead of drawing attention to it. "Just checking in" quietly asks them to account for the gap, and people avoid messages that require an apology before they require a decision.

Practical moves that change how a nudge lands:

• Attribute the nudge to the process, not to your patience. "This is the two day reminder on the homepage approval" is not a judgment about them. It is a system doing its job.
• Restate the entire ask every time. Never write "see below" or "per my last email." If the client has to scroll or search, you have added work to a message whose only purpose is to remove work.
• Shrink the ask on each pass. If they cannot decide the whole layout, ask whether they want to decide it on a ten minute call instead. A smaller decision is still forward motion.
• Put the consequence in dates, not in feelings. "Every day this sits moves launch from the 14th" is a fact about a calendar. "We are blocked waiting on you" is a complaint, and complaints invite defensiveness.
• Send when they can act, and find out when that is. Watch your own accounts for a pattern. If a client's answers keep arriving early in the week and their silence keeps landing on late Friday sends, move the ask into the window that gets replies and stop feeding the one that does not. Test it for a month before you trust it.

A 48 hour cadence you do not have to remember

Start at 48 hours, change what you ask for on each touch, and stop after four touches by escalating to a call or invoking your stated default.

• Hour 0. The request, structured as above.
• Hour 48. Same ask, fully restated, one short paragraph. Framed as the automatic reminder.
• Day 4. Shrink it. Offer the default explicitly and confirm it is reversible. "I will proceed with A on Thursday unless you say otherwise, and we can flip it at no cost until the 19th."
• Day 6. Change the channel. Two sentences by text or a ninety second phone call. Different medium, same one question.
• Day 8. Stop nudging and write the record. Item moved to on hold, here is what is now at risk, here is what restarting costs. Then move your team onto other work, and tell them why.

Two days is a starting default, not a researched optimum, and it is worth adjusting once you have watched a particular client for a few cycles. The reasoning behind it is simple: two days lands on a different day of the week and a different mood than the original, while the request is still recognizable. A daily nudge can read as pressure to the kind of client who only gets to administrative work in the evenings, after the shop is closed. A weekly nudge risks arriving after the thing has fallen out of memory and has to be re-explained from scratch. Whichever interval you settle on, it belongs in the client communication plan rather than in your head, because a cadence is only useful if it still runs during a bad week.

Build a waiting-on-client list so follow-up does not depend on memory

Keep one list of everything you are waiting on from clients, with a date, a named person, and what stops if it does not arrive, and read it at a fixed time twice a week.

Six columns are enough: client, item, what it unblocks, date asked, date of last nudge, default if no answer. One rule governs the list: nothing goes on it without a default. An item with no default is an item you will own forever, because there is no condition under which it leaves the list.

This is the piece we automated in Client Atrium, our own white-label client portal: every open approval appears on a waiting-on-client list, and a reminder goes out on its own after 48 quiet hours. It is a disclosed option, not the point. A spreadsheet and two fifteen minute calendar blocks do the same job. What matters is that one place holds the state, so the follow-up survives the week you get sick.

Some clients will never log into anything, so the mockup you sent Tuesday and are still waiting on the following Monday has to be answerable from the email it arrived in, a tradeoff worked through in the email versus portal comparison.

Start here, even if you buy nothing

Open your sent folder and find the three oldest client asks that never got answered. Rewrite each one as a single decision with two options, a date, and a default, and send them today. Then make the list, even if it is a text file, and put fifteen minutes on Monday and Thursday mornings called "waiting on clients."

One more thing, for the next proposal you write: name the approver by title in the scope, and add a line stating that if written approval is not received within five business days, the project proceeds with the recommended option and revisions after that point are billable. In our experience that clause gets far less pushback at signing than it does mid-project, and it converts your chasing from a personal favor into a term both of you agreed to.