Headshot day for 200 employees: one frame each

A 200 person headshot day is closer to event logistics than to portrait work. There is a line outside the conference room, a coordinator with a spreadsheet, and by four o'clock you have well over a thousand frames that all look like a person against the same background. This is how you get from that folder to 200 correct files.

Two hundred people is 200 shoots, not one

Run the numbers before you touch the card. Six to eight frames per person across 200 people is 1,200 to 1,600 files, a small folder by wedding standards, where a day runs 2,000 to 4,000 frames. It is a much harder folder, because a wedding is one client and one story, and this is 200 separate contests with 200 different people waiting on the answer.

The keep rate tells you the shape of the job. You are handing back roughly one file in seven, and unlike a wedding gallery there is no room to be generous: 200 people means exactly 200 finals, which is a much stricter target than a range. The general version of that arithmetic is in what a keep rate actually tells you.

The broader craft of corporate selection, the approved look, the expression register, the grid check across the whole team, is covered in our guide to culling corporate photography. Everything in there applies here. This article is about what changes when the number reaches 200: the roster becomes the deliverable, the approval loop becomes the bottleneck, and a misfiled name costs more than a mediocre expression.

That range is a decision you make at the camera, and you cannot fix it at the desk. Twenty frames per person is 4,000 files and a twenty way comparison repeated 200 times. The insurance that actually works is a glance at the back of the camera while the person is still in the chair, which is the argument in the safety shot habit.

Count people, not files

The single most useful sentence on a large headshot day is this one: a folder with 200 images can still contain two portraits of Alex and none of Priya. File counts feel like progress and they prove nothing. The only count that matters is the number of names on the roster with a picked frame beside them.

So the roster is not an attendance list, it is the spec. Ask the coordinator for it as a spreadsheet before the day, and add four columns to it yourself:

• Photographed. Did this person sit down at all.
• Picked. Is there a marked frame against this name.
• Approved. Has whoever gets to approve it said yes.
• Status. Absent, reshoot requested, name not yet confirmed, or done.

Every row must end in a state, including the empty ones. A day is not finished when the frames are picked. It is finished when no cell is blank, and the difference between those two moments is usually the week you spend chasing eleven people.

Fill the columns as you go. Checking between sitters costs five seconds and catches the person who blinked through all six frames while they are still in the building.

The failure mode is a name, not a photograph

On a normal portrait job the worst outcome is a weak frame. Here the worst outcome is a beautiful frame filed under the wrong person, sent to the whole company in a new starter announcement. Identity matching, not aesthetic judgment, is the actual problem you are solving on a 200 person day.

Build the identity into the capture sequence so it lives inside the images rather than in your memory:

1. Check the person against the roster and photograph their name card or badge as the first frame of their set.
2. Shoot the planned expression and pose variations.
3. Ask them to confirm the displayed name before they stand up, because the roster spelling is wrong more often than anyone expects.
4. Add a color label there and then for anything unusual: a reshoot, an accessibility adjustment, an executive who needs a second pose.
5. Move the name card out of frame before the next person sits.

Voice memos work too, but a photographed card puts the marker inside the sequence where a later pass cannot lose it. If a card system is not possible, shoot a slate after every tenth person. A broken identity chain is easy to repair inside a block of ten and miserable to repair across an afternoon.

Where you can, match on an ID rather than a name, because employee numbers are unique and names are not. Two Sarah Bennetts in a 200 person company is not a rare event. School photographers solved this years ago with scannable cards, and the mechanics carry across almost unchanged in the volume school photography workflow.

A technical pass, then 200 small contests

Do not start by looking for the best smile. Start by setting aside everything that could not be delivered under any circumstances: closed eyes, mid sentence mouths, a flash that did not fire, camera shake, someone walking into frame, and the test frames from the start of each lighting change. This pass is fast because it needs no taste, only a rule.

The technical bar is higher here than in most genres. Judge the nearest eye at 100 percent rather than trusting a thumbnail, and watch the frame edges: a collar folded under, a chair back in the corner, a background seam behind a shoulder.

Then work person by person, never front to back through the folder. Put one person's surviving frames on screen together and ask four questions:

Pick a deliberate alternate while you are there, usually a slightly more serious or slightly warmer version of the same person. It costs about four seconds now and it is what saves you when somebody says they dislike theirs in October.

Automatic tools earn their keep in exactly this shape of work: face grouping splits the folder into 200 stacks, and grouping near identical frames turns six candidates into one comparison. What no tool can do is confirm that the stack labeled 47 is Priya. That division, covered in what AI should never choose, is unusually clean on a headshot day. The machine sorts, you decide who is in the photograph.

The approval loop is where the day is actually lost

Picking 200 frames takes an evening. Getting 200 people to agree with you can take a month, and it is the part almost nobody prices. Decide the route before the shoot and put it in the booking email in one sentence, because after the shoot it is a negotiation.

If the company insists that everyone chooses their own, three things make it survivable. Send a narrow set, two or three frames that you have already approved technically, never the full six. Give a real deadline. And write the default into the same message: if we do not hear from you by the 20th, the marked frame is the one that goes to the team page. Without that sentence you will be chasing people in November.

Name one person on the client side who can settle exceptions, and make it clear that this is a different job from the person who booked you. A coordinator can confirm names. Only the approver decides that the head of legal gets a second pose, or that the frame with the visible lanyard has to go.

Reshoots and absences need their own list from the start

No day at this size goes cleanly. People miss their slot, arrive after the lights are down, ask for a second attempt after seeing the back of the camera, or are simply working from home that week. Selection gets much easier when those cases live in their own collection from the first hour instead of being discovered during export.

Call it something obvious, note the reason, and note who owns the fix. Then hand the client a status list in plain words rather than a folder they have to audit: photographed and picked, photographed but awaiting a retake, absent, or waiting on a name confirmation. That list is operational information for a coordinator, not something to bury in image metadata.

Two rules protect you here. Do not substitute an old staff photo to fill an empty row, however tempting it is for the coordinator, unless someone puts the approval in writing: image age, brand consistency and the person's own consent all sit behind that decision. And book the retake block before you leave the building, because a return visit for eleven people negotiated three weeks later is a different, worse conversation.

Everything you did not pick stays exactly where it is. On a corporate job that is closer to an obligation than a preference: the person who disliked theirs on Tuesday may want the other frame on Friday.

Name the files the way HR will use them

Your picks are not delivered to the person who made them. They go to a marketing coordinator who will drag them one by one into a template, and to an HR system that may want the employee number in the filename. Every rename they have to do is a small tax on rehiring you, so the naming is part of the pick.

Rename from the agreed convention, something like Bennett_Sarah_Headshot_2026, and do it after the identity check rather than before. Renaming first feels efficient and it bakes in any mistake permanently. Surname first sorts the folder the way a directory reads. If your naming is ad hoc across jobs, our file naming system guide is twenty minutes well spent.

Then split the delivery the way it will be used: full resolution masters in one folder, the web crops in another, and any square or transparent background versions in a third. Embed copyright and usage terms. Do not export your internal notes, since a note that says reshoot requested, unhappy with teeth should not travel with the file.

Keep the alternates and the originals in the client archive with a stated retention period, and tell the client it exists. When the request arrives in six months asking whether there is another frame of the sales director, an alternate you already picked turns a reshoot into a two minute reply.

Choosing a tool for this shape of day

A 200 person day punishes software built around one long timeline. What you want is grouping by face, a fast way to compare a handful of frames of one person at full size, and labels that survive the trip from the shoot floor to export. Executives and anyone with a speaking bio may need two selected poses, so the tool also has to let a person have more than one answer.

The honest test is not a feature list. Take a headshot day you already delivered, run a free trial on it, and check whether the tool's top frame for each person matches the one you actually sent. If it agrees with you on most people it will save you a real evening.

Kepla is being built for this exact comparison, a short run of nearly identical frames of one person, and for doing that first pass on an iPhone or iPad rather than waiting until you are back at the desk. It picks and hands you a shortlist, and it never deletes, moves or renames a file: every frame it did not pick stays where it is, one tap away. The picking app is still in development for iPhone, iPad and Mac. The booking page is live and free today. The same near identical comparison turns up on a very different table in picking between near identical product frames.

FAQ