Archive Treatment Work So Reusable Methods Stay Separate From Client Material
Archive Treatment Work So Reusable Methods Stay Separate From Client Material
A finished treatment leaves two things behind. One is a way of working — something you learned about how a reader moves through a sequence of images, or how a caption does its job, or where a treatment loses people. The other is a record of one job: the copy that ran, the references that were collected, the route that got cut. Both are worth something. Only one of them is portable.
The sorting rule that does most of the work here is easy to state and easy to skip: file each item by what it can actually support, not by how generic it looks once you've altered it. A client page with the logo cropped out is still the client's page. A rejected idea renamed "our method" is still the route you developed for that project. A blank component your team built may or may not be yours to reuse, depending on an agreement you should reread rather than assume.
The example below is invented. A small studio, a fictional bottled-drink client called Kaldo, one finished pitch. Nothing in it is a real archive, and none of its sorting decisions amount to clearance. Real questions about retention and reuse belong with the responsible owner and a qualified reviewer. What the example can show is the shape of the decision.
Name the transferable method without importing the project
The studio pitched Kaldo a campaign treatment built around a sequence of eight reference photographs. The first draft captioned them the way most decks do: "warm and unhurried," "intimate," "a little wistful," "quiet confidence." The client's marketing lead read the whole thing and asked a reasonable question — were shots four and seven the same setup seen twice, or two different setups? The captions couldn't answer. "Warm" describes any of the eight. It certifies each image on its own without saying how one stands to the next.
The second draft captioned the camera relationship. Shot four: 35mm at table height, hands and cup in frame, cup edge cropped. Shot seven: same setup, camera pulled back to 50mm, full table visible. With those two lines, the lead could see that the sequence repeated one arrangement at increasing distance, and could say whether that repetition was the point or an accident. The reading problem — a reader who couldn't tell escalation from variety — was solved by describing the relation between camera and subject rather than the feeling the picture gave off.
That is the transferable method: when a reference is doing structural work in a treatment, caption the camera relationship, because that is what a reader needs in order to compare one shot with the next. It comes with its reason attached. Everything else — the drink, the kitchen, those eight photographs, the arrangement they were pasted into — is the occasion.
This is where archives usually go wrong. The team decides the caption lesson is "theirs," and then files the whole deck as "the example," because the deck is where the lesson lived. But a method note doesn't need the source. It needs a demonstration of the same operation. The two are separable in principle long before anyone in the room has the standing to separate them in practice, and that gap is what the rest of the filing is for.
Separate collections by what they can actually support
The Kaldo folder holds four kinds of thing, and they do not have the same future.
Kaldo's approved copy. The studio wrote it — that authorship is real and it matters less than most people assume. It was written for Kaldo, inside a Kaldo engagement, and the same is true of the arrangement it sits in. The file held is the delivered version: the second draft, the one whose captions describe the camera relationship. The first draft's mood captioning survives as draft history, not as a page in the folder. Restricted package. Source: Kaldo. Contributor: the studio. Known permitted purpose: the project, and nothing else has been established. Unresolved: whatever the agreement says about reuse, which is a question to look up rather than recall.
Eight third-party reference photographs. Each has a photographer or a library behind it and terms of its own. Note them one at a time. "Reference images" is not a provenance entry; it's a shrug with a category around it. Known permitted purpose: internal reference during the project, as far as anyone currently knows. Unresolved: whether the licenses allowed more.
The discarded reveal. The treatment had planned to end by pulling back from the bottle on a table until a second, empty chair came into frame. It was cut before the pitch. This is a distinctive creative route — its effect depends on the sequence that sets it up — and it was developed for Kaldo. Renaming it "our reveal method" does not move it. Calling a rejected idea a method is a labeling act, and labels don't change who made a thing or under what arrangement. Even if the studio owned it outright, dropping Kaldo's reveal into another client's deck would be a separate decision from whether the archive may keep it. Restricted, with an ownership question marked open.
The blank caption component. This one is different in kind. It's a structure: a small field layout with places for setup, lens, distance, and relation-to-previous-shot. It contains no Kaldo content. It is the studio's candidate for reuse — candidate being the operative word. Whether the engagement assigned tools and templates to the client along with the deliverables, or handled them separately, is exactly the clause to check rather than assume. So it goes into the reusable collection carrying an unresolved ownership flag, and the flag is not decorative: nothing leaves that collection for a new project until the question is answered.
Two habits keep this honest. First, give every retained item a provenance entry with four parts — source, contributor, known permitted purpose, unresolved questions — because "we don't know of a restriction" is a different statement from "there is no restriction," and a future reader needs to see which one you meant. Second, treat the archive's own categories as claims about support rather than grants of right. An unlabeled copied page does not become infrastructure because it looks generic after the logo comes off. A convenient folder cannot confer a permission that no one gave.
Create a new demonstration when the source cannot travel
Now the studio wants the caption method on its shelf. It can't use the Kaldo page, and it shouldn't want to — the page in the archive is the delivered deck, its captions already repaired. What the lesson turns on is a comparison between the mood caption and the camera caption for one image, and the delivered page carries only the second of the two. So the studio builds a new example from its own material.
It photographs its own tabletop: its own cups, its own linen, its own light. The new problem is invented too — a pitch for a fictional bakery's subscription box, where the sequence follows a box from sealed to opened. The note shows the same image twice, with two captions. One says cozy, handmade. The other says 50mm, overhead, box closed, hands at the tape. Then a second image captioned same setup, 35mm, box open, contents in frame, and a sentence telling the reader what they can now do: tell that these are one arrangement at two moments, and judge whether that's the sequence they want.
Compare that with the shortcut. The tempting move is to crop Kaldo's logo off the old page and drop the image into the note. Put the two versions side by side and the shortcut fails twice. It is still the client's page — same arrangement, same third-party photographs, same copy, now with a hole in the corner where the mark used to be. And it is a weaker demonstration, because the fragment arrives with one caption and nothing to hold it against: the reader sees the repaired line without the line it replaced. Cropping changes what you see. It does not change what the item is, who made it, or what any agreement says about it.
The real risk in rebuilding is the opposite of the shortcut: abstracting so hard that the operation disappears. "Captions should be specific" is not the same lesson as "captions should describe the relation between camera and subject," and it doesn't survive contact with a reader. Specificity alone will not tell anyone how shot four stands to shot seven. What made the original lesson useful was a comparison between two captions for one image, and that comparison is what the new note has to keep — the same operation, built from material the studio owns and can authorize.
Make later reuse answerable
A colleague opening the archive six months from now should be able to tell four kinds of holding apart without asking anyone. A studio-built structure may be cleared for reuse, usable under stated conditions; or it may still be a candidate, held back by an unresolved ownership flag, which is the state the blank caption component sits in. A reference-only item, such as the third-party photographs, is there to be looked at and not copied. And a restricted client artifact — the approved copy, the cut reveal — sits still until a decision happens. Those four states should not look alike in the filing.
Each retained item should also point back to the relevant agreement or owner where that link is itself permitted to exist, so that the question "can we use this?" has a route to an answer instead of a shrug. And each should state what remains open before reuse — as a field, not a vibe. Reuse blocked pending ownership check. Owner: the client-side contact on the engagement. Note added, dated. That sentence does more for the studio than any amount of careful filing without it.
It is worth being precise about what this organization is not. It is not a legal retention schedule; it doesn't say how long anything may stay. It is not permission to publish former client work, and it is not a portfolio-credit checklist. It doesn't establish that every archived item may remain indefinitely, or that an expired project has become a free one. Possession, authorship, and permission are three different facts about the same file, and the archive's job is to record which of them you actually have.
The test of the whole arrangement is whether the method note can survive the source. If a lesson can only be taught while the client's file stays in the folder, then the lesson is still hosted by the project, and the first time someone asks for the file to be removed — or the agreement runs out, or the client moves on — the teaching goes with it. Handle it the other way and the two separate cleanly. The note names an operation, shows it on an example you own, and reaches the old work only through a record that says who holds it and what still needs deciding. Then deleting the source doesn't delete the lesson, and keeping the lesson doesn't quietly carry a client's work into someone else's pitch.
Frequently asked questions
What is the basic sorting rule for archiving treatment work?
File each item by what it can actually support, not by how generic it looks after alteration. A client page with the logo cropped out is still the client's page, and a rejected idea renamed as a method is still the route developed for that project.
How do you keep a transferable method from importing the project?
State the operation and its reason — for example, caption the camera relationship when a reference does structural work — and demonstrate it with material you own. The method needs a demonstration of the same operation, not the original source.
What should each retained item record?
A provenance entry with source, contributor, known permitted purpose, and unresolved questions. It should also carry unresolved reuse fields with the owner and date, so “we don't know of a restriction” is distinct from “there is no restriction.”
Why not crop the logo and use the old page as the method example?
Cropping changes what you see, not what the item is, who made it, or what agreements say about it. It is also a weaker demonstration because the fragment may show only the repaired caption and not the caption it replaced.
What are the four archive holding states described?
A studio-built structure may be cleared or usable under stated conditions; a candidate may be held back by an unresolved ownership flag; a reference-only item is there to be looked at and not copied; and a restricted client artifact stays still until a decision happens. These states should not look alike in the filing.