What one hour of culling actually costs

An hour of picking photos never sends you an invoice, which is why it is the least examined hour in your week. This is how to put a defensible price on one of yours: what the hour displaces, what is actually inside it, how to measure it across your next five jobs, and what buying it back would cost at this month's prices.

The hour is the unit nobody prices

An hour spent picking photos leaves almost no trace. There is no invoice, no line in the accounts, no entry in the calendar you keep for clients. It happens after dinner, and the only evidence it existed is that you went to bed later than you meant to.

That is exactly why it survives. Software announces itself once a month with a number attached, and your stack priced per wedding turns that number into something you can argue with. Your own hours arrive with nothing attached at all, so they get spent first and counted never.

This piece does one narrow thing: it puts a defensible price on a single hour of selection. Not the season, not the year. One hour. Once that number exists, every decision downstream of it, whether to buy a tool, whether to hire, whether to reprice the work, becomes a comparison instead of a Sunday night mood.

Two neighboring articles, so you know what lives where. The hours themselves, what they are worth in total, and what the work does to your judgment by frame 2,000, are set out in the true cost of manual culling. What the lag costs you separately, in late balances, cooling referrals, declined bookings and a falling rate, is worked through in the hidden cost of slow culling. This one is the unit both of those are built out of, and it is the only one you can measure tonight.

An hour costs whatever else was standing in line for it

The price of an hour is not a property of the hour. It is a property of the thing it displaced.

So price it by naming that thing out loud, in one sentence, for the specific hour you are about to spend. This is the hour I would have used to answer Saturday's inquiry. This is the hour I would have spent designing the album, which carries a real margin. This is the hour I would have slept before a 6am call time. Three hours, three completely different prices, and only one of them is anywhere near zero.

This is also why the same four hour first pass feels harmless in February and ruinous in June. In February it competes with a television program. In June it competes with an inquiry that expects a reply the same day, and with the recovery you need before the next shoot. The work is identical, the cost is not, and any calculation that gives you one flat figure for the whole year is hiding that.

For scale on what a peak season hour is up against, The Knot's 2025 Real Weddings Study, drawn from 10,474 US couples married in 2025, puts the average US wedding photographer at $3,000, with quartiles at $1,500, $2,800 and $4,700. The arithmetic on bookings you actually turned down belongs in the hidden cost of slow culling and there is no sense doing it twice. The narrower point here is that an hour in the middle of your season is competing with a fraction of a number that size, and an hour in the middle of winter is competing with the sofa.

What is actually inside the hour

Treating culling as one block is what makes it hard to price and impossible to improve. Open it up and there are four distinct jobs inside, and they are not worth the same.

The expensive minutes are almost never the ones you would describe as culling if somebody asked. They are the repetitive comparison minutes: which of the six flash frames caught the expression, whether frame nine or frame eleven has the sharper eye. That work is slow, it is tiring, and at the end of it nobody can tell which one you chose.

The narrative minutes are the opposite. They are fast, they are the reason the gallery looks like yours and not like anyone else's, and they are the last thing you should be trying to speed up. A sensible efficiency project takes screen time out of the first two rows of that table and leaves the bottom two completely alone. If it makes the bottom two faster, it is not an efficiency project, it is a quality cut.

Measure your own hour rather than borrowing anyone's

You cannot price an hour you have never timed, and memory after a wedding weekend is not evidence. Take five recent jobs and record five lines for each. A notes app is fine. The point is that the numbers come from a stopwatch rather than from a feeling.

That last line matters more than people expect. Twenty minutes waiting on previews is twenty minutes of production time, and it belongs in the cost even though it does not feel like culling. Leaving it out is how a four hour job gets remembered as a two hour one.

Run those five jobs and you will probably find frame count is a weaker predictor than you assumed. A bright 2,500 frame daylight wedding moves quickly. A 4,000 frame reception with flash, two shooters and long bursts can take twice as long per thousand frames. Family formals are quick technically and slow emotionally, because you are choosing the one frame where nine people are all doing the right thing at once. Industry norms put a full wedding day at roughly 2,000 to 4,000 frames captured and 400 to 800 delivered, but ranges like those are for planning, not for pricing your evening. If you want to see how the shape changes by job, culling time by shoot size breaks it down, and how long culling takes covers the question in general.

What you are after is one sentence you can say with confidence: my typical wedding runs about this many frames and takes about this long to reach a shortlist. Without that baseline you will end up judging every change on how it felt rather than on what it did.

What an hour buys, and what buying it back costs

Now the comparison the exercise is for. You have an hour with a price on it. What would it cost to not spend it?

Take the going rate for selection software, from each vendor's own page this month. Aftershoot Select is $10 a month billed annually, $120 for the year, unlimited images, desktop. Narrative Select Lite is also $10 a month on annual billing, with Standard at $20. Imagen's culling add-on is $12 a month billed annually, or $18 monthly, on top of an editing plan, and it is cloud based, so the photos upload first. Prices checked August 2026.

Hold those against one of your hours. If your hour is worth $85, a year of the cheapest AI selection plan costs you about an hour and a half of your own time. If your hour is worth $40, it costs three. That is the exchange rate, and it is the only reason the arithmetic on these tools is rarely close. The full version of that memo, with the three defensible ways to value your hour and the traps underneath the division, is in the business case for culling software.

Two things that comparison does not settle. The first is which pricing model suits your volume, because a flat fee and a per image rate behave very differently across a lumpy season, which is worked out in flat rate versus per image pricing. The second is whether a person should do it instead, which has its own math in AI culling versus outsourcing. Get a real quote for your own volume there rather than trusting a figure in an article.

Whatever you compare, compare the same thing: the effective cost of one finished selection, not the headline on the invoice. A cheap option that hands back a set you re-check for an hour is not cheap. It moved the hour. If you want your whole software line costed instead, the cost calculator does that side.

When the hour you saved was not saved

Speed is only savings when the work stays done. Four ways it quietly does not.

The rushed pass that comes back. Flying through the first pass and missing the grandmother's reaction costs more than the minutes it saved, because finding it later means a catalog search, a second export and an apologetic email.

The unresolved maybe. A maybe flag is a decision you have agreed to make twice. A few are fine. Two hundred of them is a second cull you scheduled without noticing, and it is the most common reason an evening turns into a weekend.

The wrong bottleneck. If your real problem is a slow card reader or previews that build while you wait, no selection tool will fix it. Time the stages separately first, or you will pay to speed up the part that was already fast.

The learning curve nobody counted. A new tool that takes a month to trust may still be worth it, but the month is a real cost and it belongs in the first year's total.

One caveat worth stating plainly: an hour saved is not automatically an hour earned. It turns into money only if you fill it with paid work, or with rest that lets you take a job you would otherwise have declined. If it becomes an hour of scrolling, it bought you nothing in cash, which is a fine outcome as long as you know which one you bought.

What to change before you spend anything

Most photographers can take a bite out of the hour without buying anything. Six changes, in the order they usually pay off.

1. Decide the target delivery count before you open the folder. Somewhere in the 400 to 800 range is normal for a full wedding day. Picking without a target is how a first pass drifts.
2. Make one fast pass that only throws out the obvious failures. No zooming, no comparing, no second guessing.
3. Compare bursts only after that pass, when there are fewer of them left to compare.
4. Use the maybe flag sparingly, and clear every one before you stop for the night.
5. Do the story pass separately from the technical one. Mixing them is what turns one sitting into three.
6. Write down the result. Without a before, you cannot claim the after.

Then look at what a tool would take off the top. The honest promise of AI selection is that it handles the checks: focus, open eyes, near identical frames. It does not know which frame carries the day. Nothing does, except you.

Where Kepla fits

Kepla clears the obvious misses and hands you the clean frames, so the first pass can happen on the drive home rather than on Sunday night. It never deletes, moves or renames a single file, and every frame it did not pick stays one tap away. The picking app for iPhone, iPad and Mac is still being built. The booking page is live and free today.

FAQ