Cull before you outsource: cleaner selects, smaller bills

Sending an editor the whole card feels efficient until the invoice arrives and the return folder holds six versions of the same moment. An editing partner can make a photography business scalable, but nobody can guess your must-have expressions or family dynamics from a folder of thousands of candidates. A clean selection packet is how you buy their speed without paying them to make your decisions.

Decide what you are actually hiring them to do

There are two very different services hiding under the phrase outsourced editing. A cull and edit service chooses the images and then applies your corrections. An edit only service works from the selection you hand over. The first one can rescue you in a backlog, but it needs a detailed brief and it costs more, because somebody has to look at every frame you shot.

That is fair pricing, not a rip-off. Reviewing thousands of captures is real labor whoever does it. The mistake is assuming that sending a card dump is free because the per-image edit rate looked low.

For most established wedding workflows the answer is to keep the selection in house and outsource the repeatable finishing work. Industry norms put a full wedding day somewhere in the range of 2,000 to 4,000 RAW frames with 400 to 800 delivered, and your editor should be receiving something close to the second number, not the first. That lets them spend their attention on exposure, white balance, crop consistency and your look, rather than deciding which smile was the real one. If you have not worked out where your own ratio sits, that is worth doing first, and we broke it down in shot to delivered ratio.

Choosing between an editor, AI editing software and doing it yourself is a separate decision with its own trade-offs, and we walked through all three in outsourcing your editing. This piece assumes you have picked a person and want the job to cost less and come back cleaner.

The count you send sets the bill, not the rate you negotiated

Photographers shop for editors by comparing per-image rates. That is the smaller half of the arithmetic. What actually moves the invoice is how many files you upload, because the count multiplies whatever the rate turns out to be.

Run your own numbers before the next job goes up. There are only three inputs, and you already know two of them.

The gap between the first two rows is usually larger than photographers expect, and it is worth knowing exactly, in your own currency, before an emergency makes the decision for you. The comparison between paying a human for that first pass and handing it to software has its own numbers, and they sit in AI culling versus outsourcing the cull.

None of this means you should never buy a cull. It makes sense in a genuine backlog, when your style is documented and very repeatable, or when you are still shooting the next job. The point is to choose it deliberately and to put its true cost inside your package pricing, rather than discovering it in October.

Build the same three-pass packet every time

A selection packet is not a folder, it is a decision that has been made and recorded. Use the same three stages on every job so that the packet looks identical whoever is receiving it.

Technical pass. Set aside blinks, focus misses, test frames and accidental captures. Keep a full burst only where the action genuinely needs it. This is the mechanical stage, and it is the one that responds best to software.

Story pass. Pick the strongest frame from each beat: getting ready, entrances, vows, family, portraits, reception, exit. Keep a supporting frame only when it adds a different reaction or a different angle, not because it is also good.

Editor pass. Apply the rating your editor will filter on. A convention that survives contact with real jobs: three stars for a standard edit, four stars for a hero image that you will look at yourself, and a color label for black and white candidates or anything needing retouching. Write it down once and never change it mid-season.

In Lightroom Classic, filter to your three and four star files before you export a Smart Preview catalog or share the folder. If your first pass happens in Photo Mechanic, write the stars to XMP before import so there is exactly one source of truth about what was selected. Whether you mark keepers or mark failures changes how long that first stage takes, and the answer flips with the size of the job, which is the argument in cull in versus cull out.

Make the brief impossible to misread

A folder of selects with no instructions is an invitation to guess, and every guess is a revision waiting to happen. Send a one page job sheet with the files, every time, in the same order.

It needs the delivery target, your crop rules, a reference gallery you still stand behind, the lighting exceptions on this particular job, the export specification, and a list of the people and moments that must appear. Names matter more than photographers expect. The editor does not know which of the four grandmothers is the one who traveled from Lisbon.

Write in observable terms rather than adjectives. Warm and romantic is not an instruction. Keep reception skin warm but do not lift the blue uplighting is an instruction. Image 842 needs the boutonniere straightened, do not swap it for a similar pose is an instruction. A note a person can check their own work against will save you a round trip.

Anchor images do the rest of the work. Twenty to thirty frames covering window light, flash, tungsten, sunset, dance floor and any black and white intent will teach your look faster than any paragraph. That technique works the same way for a human editor and for an AI profile, and it is covered properly in our piece on handing off your editing, so it is enough here to say that the anchors belong in the packet alongside the brief.

Count revisions as part of the price

The quoted rate is the visible cost. The revision loop is the one that quietly eats the margin, because it is paid in your evenings rather than on the invoice, which is exactly why nobody tracks it.

Start tracking it. After each delivery, note why every correction happened, in one of four buckets: wrong selection, color mismatch, crop, or client request. Four buckets, thirty seconds, one line per job.

The pattern tells you where the fix belongs. Color and crop notes mean your anchors or your brief need work. Client requests are the cost of doing business and should stay. But if selection keeps appearing, your editor is not the problem: they are working on frames that should never have been in the folder, or they are missing frames that never got flagged. A thirty minute review of returned work becomes two hours when the set is full of near-identical alternates you have to compare all over again, which means you have now paid twice to make the same decision.

The goal is not zero revisions. It is to reserve revisions for genuine finishing choices, so that the ones you do send are about taste rather than about cleanup.

Decide up front which images keep you in the loop

Not every frame deserves the same approval process, and treating them all as equal is how photographers end up reworking a third of the gallery after it comes back.

Split the job before it goes out. Standard documentary coverage needs exposure, white balance, crop and your baseline treatment, and it can come back finished. Hero portraits, family formals and anything heading into an album deserve your own final look. Flag those with the four star rating or the color label from your packet convention so they do not disappear into the standard batch.

That gives you a hybrid: outsource the repeatable majority, spend your own editing time on the minority where your eye is worth the most. It also protects the working relationship, because an editor cannot be blamed for not applying a level of care nobody asked them for.

If you find yourself reworking a hundred and fifty images after every return, do not add another round of notes. Look at whether those images should have been in the hero category from the start.

The five minute check before you upload

Think of the handoff as an instruction set rather than a file transfer. Almost every expensive mistake in outsourced editing is a packaging error, and packaging errors are cheap to catch and painful to discover at revision time.

The complete package is short: the selected files only, marked with one consistent rating; the delivery target, stated as a number, with a note about whether alternates are wanted; twenty to thirty anchor images covering each lighting condition; notes on crops, black and white candidates and any image-specific fixes; the must-deliver list of people and moments; and the export specification, meaning color space, file size, naming convention and the due date.

Then verify it. If you are sending a Smart Preview catalog, open it yourself on a different drive or machine first. If your editor works from RAW files in a shared folder, confirm the folder actually contains every selected file rather than previews of them. Count the files against your expected delivery number. Five minutes of that beats finding out at revision time that the ceremony is missing.

Where a machine first pass fits into this

Assisted culling does not replace an editor. It changes how long it takes you to build the packet, which is the part of this whole process that sits between you and the handoff, usually at an hour of the evening when your judgment is at its worst.

The mechanical stage is what software is genuinely good at: sharp or not, eyes open or not, the fourth frame of the same burst. Your story pass and your hero flags are the parts a machine cannot do for you, and they are also the parts your editor is not being paid to guess. Whether that trade is worth a subscription for your particular workflow is a measurable question, not a matter of opinion, and there is an audit for it in when AI culling actually saves time. What an hour saved is worth in your year is arithmetic we already ran.

That first pass is what Kepla is being built to do, and here is the honest status: the picking app for iPhone, iPad and Mac is still in development. The plan is a first pass on the phone on the drive home, then the same shoot picked up on the Mac when frames need a big screen, and nothing is ever deleted, moved or renamed, so every frame it set aside stays one tap away. The booking page is live and free today.

FAQ