The delivery spec that earns repeat UGC orders
Most creators lose the re-order at handoff, not on creative. The aspect-ratio set, safe zones, caption files, frame rate and naming that make a folder usable.
The video that gets re-ordered is rarely the best video in the batch. It is the one the media buyer could upload without opening an editor.
That sounds cynical until you watch the other side of the handoff. A buyer receives a folder, needs a 9:16 for TikTok and a 1:1 for the feed by end of day, and finds one 16:9 master with burned captions sitting under the platform's own UI. Now they have to crop it themselves, the captions get clipped, and the job that was supposed to save them time cost them an hour. They will still pay the invoice. They will not send the next brief.
Everything below prevents that. None of it is creative work, all of it is repeatable, and it is the cheapest retention lever a UGC creator has.
Ship the aspect-ratio set, not the master
A deliverable is not a file. It is a set of files covering the placements the client actually buys.
The default set for a UGC ad:
| Ratio | Where it runs | Notes |
|---|---|---|
| 9:16 | TikTok, Reels, Shorts, Stories | Treat this as the master if the shoot was vertical |
| 1:1 | Feed placements, some in-feed ad units | The crop most likely to lose a subject at the edges |
| 16:9 | YouTube, site placements, connected TV | Only include when the client has told you they use it |
Versely's editor reformats between those three through edit_video's aspect-ratio setting, and the change video aspect ratio task is the one-step version. You describe the target to the agent rather than picking from a menu; 9:16, 1:1 and 16:9 are the documented common targets. If a client wants something outside that set — 4:5 for a particular feed placement, say — confirm it before you build the batch, not at handoff.
The mistake worth naming: cropping a wide original down to 9:16 cuts content near the left and right edges of the frame. If the product label lives at the edge of a 16:9 master, it does not exist in the vertical version. Decide the crop while you are still generating, not while you are reformatting. Which aspect ratio wins where covers the placement side of that decision, and picking the right format per platform is the practical checklist.
Build to the safe zone, then verify the export
Every platform lays its own interface over your frame — captions, handles, buttons, the CTA row — and the region that survives is different per platform and changes without announcement.
Do not memorise pixel numbers. Anything published in a blog post, including ours, is a snapshot of a spec that moves. The working discipline is:
- Keep on-screen text inside a conservative central band rather than at the extremes, so a change to the platform UI does not immediately eat it.
- Check the platform's own current spec page before a large batch, not a screenshot of it from a year ago.
- Verify the actual export on a phone before delivery, not the preview in the editor.
That last step is the one people skip and it is the one that catches the real failures. Where on-screen text actually survives platform UI goes into why there is no single number to memorise, and the vertical ad safe-zone map is the version to hand to a client who asks why their headline is behind the follow button.
Deliver a clean master and a captioned version
This is the single change that most improves how a buyer feels about your folder.
Versely's captioning burns subtitles into the video pixels. It does not output a separate toggleable .srt or .vtt sidecar track — burned-in captions are the output, timed to the speech by the add-captions task. That is fine for social, where captions need to be visible in a silent autoplay feed anyway, but it has one consequence for delivery: once a caption is burned, the client cannot change the wording, translate it, or remove it for a placement that supplies its own subtitles.
So ship both:
clean— the cut with voiceover and music, no on-screen text at all.cap— the same cut with burned captions in the agreed style.
The clean master is what makes the folder reusable. A buyer who wants to test a different headline, run the ad in a market that needs different subtitle wording, or hand the asset to a partner does it from the clean file without coming back to you. Counterintuitively, giving them that file generates more re-orders, not fewer, because the asset stays in rotation longer.
If the client needs a sidecar subtitle file for an accessibility or CMS requirement, say so in the quote — that is a separate deliverable produced outside the burn-in pass, and a video transcript is the starting point for it. Discovering it at handoff is exactly the surprise that costs the re-order.
Keep the caption look constant across a batch — that is a standard you can lock once rather than re-decide per video, which is what keeping captions on brand across every video is for.
State the frame rate, and make it one number
Frame rate is the least glamorous item on this list and the most common source of a re-render.
Versely generates video at 25 fps by default. If a client's existing library is 24 or 30, mixing rates inside one campaign produces judder on cuts and, worse, means the buyer's own editor has to conform your files before they can be dropped into an existing timeline. Neither of those is your fault and both become your problem.
Three rules:
- Ask once, at intake. "What frame rate is your existing footage?" is one line in a kickoff email and it removes an entire category of rework.
- Keep one rate across a batch. Never deliver a mixed-rate folder. If part of the batch was sourced elsewhere, conform it before delivery.
- Write the rate into the delivery note and the file name. A buyer should not have to check the file's properties to know what they received.
The frame rate entry covers why the number matters at the playback end. For delivery purposes the only thing that matters is that it is stated and consistent.
Name files so the ad account can read them
Naming is where a delivery either becomes data or stays a pile. The buyer is going to load these into an ad platform, run them against each other, and later want to know which hook won. If your names cannot answer that, the answer gets lost.
A working scheme:
brand_product_hook02_var-b_9x16_25fps_cap_v1.mp4
brand_product_hook02_var-b_9x16_25fps_clean_v1.mp4
Reading it left to right: who it is for, what it sells, which hook, which variant of that hook, the ratio, the frame rate, whether captions are burned, and the version. Every field is something a buyer will filter or sort by inside their own asset library.
Two conventions that matter more than the exact format:
- Hook number is stable forever. Hook 02 in January is the same hook concept in June. That is what makes performance data comparable across months, and it is the backbone of any hook rate analysis.
- Never encode the date as the version.
v1,v2,v3survives a re-delivery;final,final2,FINAL-USE-THISdoes not.
Deliver one folder per batch, one subfolder per hook, ratios flat inside. No zip inside a zip. No Google Drive folder that requires an access request — the number of re-orders lost to a permissions prompt on a Friday afternoon is not zero.
The handoff message
Three sentences, sent with the link:
Batch 4 is up — 6 hooks, each in 9:16 and 1:1, clean and captioned versions of every cut, all 25 fps. Naming is
brand_product_hookNN_var_ratio_fps_type_version, so hook numbers match batch 3 and you can compare them directly. Notes on this round go in one reply and I'll apply them in a single pass.
That message does three jobs: it tells them what they have, it tells them the data is comparable to last time, and it defines the revision round before they define it for you.
On revisions — the editor's preview pass renders at 480p for free, with a short per-user cooldown between passes, and the final export is charged once regardless of clip count. So a consolidated notes round is cheap to apply and re-check at preview resolution before you spend on the export. There is no free plan behind that — every generation costs credits — but the editor is the one place a revision round does not compound.
FAQ
Should I deliver the clean master even if the client did not ask for it?
Yes. It is the file that keeps the asset alive after the first caption test, and it costs you one extra export from a timeline you already built. The only reason to withhold it is if your contract prices the clean master separately as a source-file licence, which is a legitimate structure — but then say so up front rather than letting the client discover the omission.
What if a client asks for a subtitle file rather than burned captions?
Treat it as a separate deliverable and quote it. Versely's captioning burns into the pixels rather than producing a toggleable track, so a sidecar file is produced outside that pass. Flagging it during scoping takes ten seconds; discovering it at handoff costs the re-order.
Does frame rate really matter for social ads?
For playback on a phone, rarely. For the client's editor, always. The moment your file has to sit on a timeline next to footage at a different rate, someone conforms it, and that someone is either you before delivery or them after. Ask at intake and keep the batch uniform.
How many aspect ratios should a standard package include?
Two by default — 9:16 and 1:1 — with 16:9 added only when the client has told you they run placements that need it. Shipping ratios nobody buys inflates your export count without improving the delivery. Fix the naming scheme once alongside it and never renegotiate it mid-engagement; if you run the same structure weekly, a saved and reusable workflow keeps the build consistent, which is what keeps the names honest. The wider version of this is the content handoff pipeline.