Run a Client Photo Approval Round Without Email

Why the email approval round falls apart

An email approval round has four jobs: get the photos in front of the client, collect a decision on each one, keep the notes attached to the right frame, and record what was agreed. Email does the first job and fails the other three.

The failure is structural, not a matter of discipline. Attachments hit a 25 MB ceiling, so you send a link to a folder instead. The client downloads the folder, opens the files in whatever their laptop uses, and replies in prose: “Love 4, 7 and the one with the blue chair. Can we brighten the third one?” Now you are reconciling their numbering against your filenames, guessing which frame has the blue chair, and asking which “third one” they mean.

Then a second stakeholder replies-all with a different shortlist. Neither list is authoritative. You have two threads, no record of what was approved, and a client who believes the round is finished.

There is a second failure that shows up later. Email has no notion of a version. When you re-send frame 12 brightened, it arrives as a new attachment in a new message, and the client now holds two copies with no indication of which is current. A month later somebody forwards the wrong one to the printer. Every photographer who has worked this way has a story that ends with the wrong file in production, and none of them were caused by carelessness — they were caused by a medium that treats each message as unrelated to the last.

The third failure is the quietest: email keeps no state. There is nowhere to look to answer “what is still outstanding?” The only way to find out is to reread the thread, which is why the question usually gets answered by sending another email and waiting.

The real cost, in hours

Put numbers on it before you change anything. Across a month of commercial work, the admin time attached to approvals tends to break down like this — measure your own, but this is the shape it usually takes:

At eight shoots a month that is roughly fourteen hours of unbillable reconciliation work — close to two full working days spent turning prose back into filenames.

The five-step approval round

This is the process, in the order it should run. It works with any tool that keeps feedback attached to the image; the sequence matters more than the software.

1. Cull before the client sees anything

Send selects, not everything. A client shown 400 frames does not review 400 frames — they review the first thirty, get tired, and reply with something vague. Sending 40 strong frames produces a faster and better decision than sending 400 mediocre ones. Culling is your job, and doing it is part of what the client is paying for.

2. Name one decider before the link goes out

Ask directly: “Who gives the final yes on these?” Other people can comment. One person approves. Without this, you get three shortlists and no decision, and the tie-break lands on you — which is the one job that genuinely is not yours.

3. Send one link with the deadline in the first line

Not an attachment, not a folder, not a zip. One link, and the date in the opening sentence rather than buried under three paragraphs of pleasantries.

4. Let feedback land on the frame

The single highest-value property of a review link is that a comment is attached to the image it describes. “Brighten this” on frame 12 needs no translation. The same words in an email cost you a clarifying message and a day.

5. Close the round explicitly

When the decider approves, say so in writing: what was approved, how many frames, and what happens next. This is the record you will want in three months when someone asks why a frame was not retouched.

What this looks like in a message

The whole thing fits in five lines:

“Here are the 42 selects from Tuesday: [link]. Could you mark the ones you want retouched by Thursday the 14th? That keeps final delivery on the 21st. Anyone on your side can leave comments, but I will treat your yes as the final one. If I have not heard back by Friday I will proceed with my recommended set.”

Every element of the process is in there: the cull is done, the link is single, the deadline has a date and a consequence, the decider is named, and the silence default is stated before it is needed. It takes about a minute to write, and it is the difference between a round that closes on Thursday and one that is still open in two weeks.

Setting a deadline clients actually meet

Most missed approval deadlines are not rudeness. They are a missing default. The client does not know what happens if they say nothing, so saying nothing feels free.

Fix it by defining the default in advance, in the same message as the link:

• State the date, not a duration. “By Thursday the 14th” beats “within a week”, which starts an argument about when the week began.
• Attach the consequence. Late selects move the delivery date. Say that up front, once, without menace.
• Define the silence default. “If I have not heard back by Friday I will proceed with my recommended set” converts silence into a decision instead of a stall.
• Send one reminder, on the morning of the deadline. Not three. One, timed for when action is still possible.

Deadlines set this way tend to land the same day or the next. The change is not pressure — it is that the client now knows what the alternative to deciding is.

Handling contradictory feedback without another round

Two stakeholders, two opposite notes on the same frame. Handled badly this costs a full extra round. Handled well it costs one message.

Do not average the notes and do not pick a side on taste. Send the contradiction back to the decider, stated neutrally and with a recommendation attached:

“On frame 12, Sam asked for a tighter crop and Alex asked to keep the room visible. Both are reasonable — the tighter crop reads better on mobile, the wide version shows the space. My recommendation is the tight crop for social and the wide for the site. Happy to do either if you would rather pick one.”

Three things make that work. It names both people, so nobody is overruled quietly. It gives the trade-off in terms of use rather than taste. And it ends with a recommendation, so the decider can reply with one word.

The general rule: never send a contradiction back as an open question. An open question generates discussion, and discussion generates another round.

The same principle applies when the contradiction is between the client and you. If a requested edit will not work — a crop that loses the logo, a brightening pass that clips the highlights — say what breaks, offer the nearest thing that does work, and let them choose. “That crop cuts the logo; here it is as tight as I can go while keeping it” ends the exchange. “Are you sure?” extends it.

What good looks like

A healthy approval round has four properties, and you can check all of them without any particular tool:

Most of the effort should sit in culling, which is judgement work only you can do. Admin should be the thinnest slice on the chart. If reconciling filenames against prose is taking longer than choosing the photographs, the process is the problem, not the client.

Four checks, at the end of every round:

1. Every note is attached to a specific frame.
2. Exactly one person approved, and you can name them.
3. The approval is written down with a date.
4. You sent one link, and no attachments at all.

If all four hold, the round will close in days rather than weeks — and the next one will close faster still, because the client has learned what the process expects of them.