Ecommerce Video Production Workflow for Catalogs With Hundreds of Products
How I run product videos for catalogs with hundreds of SKUs now: one brief per collection, a project per launch, custom statuses per SKU, and review built into every stage.
The ecommerce video production workflow in my agency used to be basically a shared drive full of files called things like final_v3_green_REAL.mp4, a WhatsApp group where the brand's content lead sent voice notes late at night, and a spreadsheet my project manager updated by hand that was always two days behind the timeline, right.
The failure that hurt most was always the same one. A colourway would get reshot, the editor would export a new file, someone would upload it under a slightly different name, and the client would open the old version and leave notes about the wrong shade of olive, so we'd lose a whole round on a product that was already fixed, which on a catalog of a few hundred SKUs adds up really really fast.
Now I run it very differently, and most of the change is about where review sits. Every collection gets its own project, every SKU carries a custom status, a reshoot stacks on the same card as MV2, and the client drops a note on the exact frame where the zip looks wrong, so here is the whole pipeline the way I actually run it.
What an ecommerce video production workflow looks like
People usually picture the shoot when they think about product video, the turntable and the model walking toward camera, right, but in my experience the shoot is maybe a fifth of the work. My stages are brief, shot list, shoot, edit, review, channel cuts and publish, and what makes ecommerce different is that you run them two hundred times with tiny variations between SKUs.
I wrote a separate breakdown of the brand video production process, and the stages look similar there, but a brand film mostly fails when the idea is wrong, while a catalog fails through drift, where SKU 140 quietly looks different from SKU 12 because another editor cut it in another week with a slightly different grade.
So the whole video production workflow management side of things exists to stop drift and lost rounds. You see what I mean here, the glamorous part is the shoot, but the money is made or lost in how cleanly two hundred files move from edit to approved.
A worked example with one 80 SKU collection
Say a brand drops 80 SKUs and wants a product page loop, a marketplace cut and a vertical social cut for each, which is illustrative math but close to what lands on my desk. That's 80 masters plus 240 channel cuts, so 320 deliverables, and if every master takes two rounds and every channel cut gets one light check, you're at roughly 400 review passes before anything gets reshot, right.
I'd lock the brief and shot list in the first two days, shoot over two or three days because we batch by setup, take edits back in waves of about twenty, and start a SKU's channel cuts only once its master is approved, so cutdowns trail masters by a few days instead of colliding with them.
Briefs and shot lists that scale across SKUs
The biggest mistake I made early on was writing one brief per product, because by the fortieth brief the copies had started to disagree with each other, right. Now I write one master creative brief per collection covering the look, pacing, music rules, grade and channels, and each SKU gets only a short row in the shot list.
For instance, on a footwear collection the master brief says a 3-second hero spin, then a macro on the sole, then the on-foot shot. The SKU row just says the colourway, the one feature worth calling out, any claim the brand legally can't make, and which channels it's going to, so the editor can cut straight from the row without messaging anyone.
- SKU code and colourway exactly as on the product page
- One hero feature to show up close
- Claims the brand cannot make on camera
- Required channels and aspect ratios
- A reference frame from an approved SKU
That used to live in a Google Doc nobody opened after week one, and now it lives in PlayPause Playbooks, which come with Agency and up. The brand playbook holds permanent rules like the logo lockup and the grade, a folder playbook per collection holds that launch's specifics, and checklists aren't done until someone actually ticks them, which sounds small but trust me on any level it changes behaviour. Playbooks can also be shared as a no-login page, so a new freelancer reads the rules before day one.
On shoot day, shoot by setup
On set we shoot by setup rather than by SKU, so every hero spin on the turntable happens in one block, then every macro, then every on-foot shot, because relighting forty times for forty products is where shoot days quietly disappear. I also want a card with the SKU code in the first frame of every take, basically a cheap slate, so the editor never guesses whether a clip is the olive or the khaki.
I treat the first exports a bit like dailies on a film set, a quick look at five or six SKUs while the crew is still there, to catch a wrong product or a smudged lens before it's baked into eighty files.
Organising projects by collection or launch
This is where catalogs break most tools, because a catalog tends to be twenty launches a year with fifty or a hundred SKUs each, right. My rule is one project per collection, so the spring linen drop, the festive gifting edit and the restock in six new colours each get their own project, and so on. Creator stops at 30 projects while Agency has unlimited projects, so a busy ecommerce video team working this way belongs on Agency.
After the shoot a few hundred edits come back, and batch uploads drop the lot into the collection project in one go.
Each SKU gets its own card with a custom status that mirrors the pipeline, so for us it's Shot, In edit, Internal check, With client, Changes, Approved and Published. The brand's content lead can run down the statuses and see which SKUs are waiting on them, which ends a lot of awkward status calls, and with several collections live at once, managing every project in one place keeps you sane. If you're splitting a collection across several people, I also wrote about coordinating freelance editors on catalogs.
Review stages for product accuracy and brand
I split review into two passes, and this one decision probably saved more rounds in my agency than anything else, right. Pass one is purely product accuracy, so the right SKU, colour true to the physical product, a readable label and no claim legal would kill, and pass two is brand, which covers pacing, music, grade and the end card.
The catch here is that when you mix them, the client's taste opinions drown out the accuracy problems, and an accuracy problem that ships comes back as a return. So the brand's product person does pass one on the rough cut before the marketing lead ever sees it, does that make sense, right.
Review accuracy first and brand second, because a wrong shade that ships comes back as a return.
The reviewer opens one link in a browser with no account and no install, and every comment sticks to the exact frame, so the editor lands on the wrong heel stitching instead of scrubbing for it. They can draw on the frame and leave a range comment across a whole spin, and since colour bites ecommerce hardest, I went deep on catching wrong product shades separately.
Reshoots are where version stacks earn their keep. When a colourway gets reshot, the new file goes onto the same card as MV2, sitting on top of MV1, and on Agency I use side-by-side version compare to check that MV2 fixed the olive without breaking the pacing. I also watch who-watched analytics, which show who opened the link, when and from which city, because to be very honest a lot of delays I used to blame on editing were a link sitting unopened in someone's inbox, which I unpacked in the real cost of slow approvals.
Channel cuts for site, marketplace and social
Once the master cut for a SKU is approved, and only then, we make the channel cuts, because resizing a video that's about to change is paying for the same work twice, right. For most catalogs that means a silent loop for the product page, a marketplace cut that follows that marketplace's rules, and a vertical social cut with a hook in the first second.
Each channel cut gets its own card named with the SKU code and the channel, because a version stack should mean the same deliverable getting better and a 9:16 cut is a different deliverable altogether. Review there is lighter, just framing, safe zones, text size on a phone and the first second, and I tell clients that upfront so they don't reopen creative decisions.
Master and cutdowns reviewed together, so notes collide and resizing gets redone
Master approved first on its own card, then channel cuts get a light framing check
Publishing is the last stage, and a SKU only moves to Published once someone has seen it live on the product page, because a loop that looked perfect in review can still get cropped by the store theme. On Agency the brand's web team can save approved files from the review link straight to Google Drive, so nobody asks me to resend the final. If your channel cuts are really paid social ads tested every week, that's a different rhythm, covered in weekly performance creative testing.
Roles and access for freelance editors
Most ecommerce studios run a small core team plus a rotating bench of freelance editors, and access gets messy quickly, right. PlayPause has four roles, Owner, Admin, Manager and Editor, so in my setup the Owner and one Admin run the account, the studio lead is a Manager, and every freelancer comes in as an Editor.
Pricing is flat per workspace rather than per user, so adding a fifth freelancer for a big launch doesn't change the bill, and Agency covers 50 members, which is a very very deep bench. Frame.io starts at $15 per user per month, and per-user pricing makes teams ration logins, which is exactly how files drift back onto WhatsApp.
For one-off freelancers who aren't members, guest version upload on Agency lets them put a new version straight onto the card while PlayPause collects their name and email. Share links can be password-protected and revoked instantly, so the day a contract ends I revoke any link I sent that freelancer. Editors in Premiere get the Premiere Pro panel, where clicking a comment jumps the playhead to that frame, while editors on DaVinci Resolve or Final Cut export, upload, review and then work through the notes in their own editor.
Where the product video production process breaks at scale
A clean pipeline on paper is very different from a Tuesday with a launch on Friday. The first place it breaks is not locking a reference, so once the first SKU in a collection is approved I give it a custom status called Reference and every editor matches against it, otherwise by SKU 60 the collection no longer looks like one collection.
The second is feedback leaking out of the tool, where the founder sees something on Instagram and sends a voice note. My rule is that if the note isn't on a frame it doesn't get actioned, which feels harsh for about a week and then everyone just does it, right.
Voice notes and screenshots get moved onto the exact frame first, so every change has one home and one owner.
The third is naming, and I'm pretty sure this causes more lost rounds than bad editing does, so every file starts with the SKU code and then the channel, and the version number lives in the version stack, which is the cheapest habit there is for product video at scale.
Frequently asked questions
How many people does an ecommerce video team need?
For a catalog of a few hundred SKUs a year, I'd say one person who owns the brief and the playbook, a shooter or small crew, two or three editors, and one person on the brand side who is allowed to approve. The approver is the one teams forget, right, and without a single named approver review turns into a committee and every collection slows down.
Should channel cuts be versions or separate files?
Separate files, in my opinion, and I've tried it both ways. The master's revisions live in the version stack as MV1, MV2 and so on, because those are the same deliverable improving, while every channel cut gets its own card, so approving the 9:16 cut never gets confused with approving the product page loop and the statuses stay honest.
Which PlayPause plan fits catalog video work?
For most catalog teams I'd point to Agency at $19 a month or $199 a year, because it has unlimited projects, 50 members, 250 GB, side-by-side version compare, guest version upload and Playbooks. Creator at $9 suits a small brand with a handful of launches, though its 30 projects and 30-day links run out quickly, and Enterprise at $27 adds 150 members, 1 TB, white-label and a custom share domain.
Can clients review product videos without an account?
They can, and that's the main reason review actually happens on time in my agency. The client opens one link in a browser with no account and no install, comments land on the exact frame, and they can draw on the frame to circle a detail. If a link needs to stay private before launch, I add a password and revoke it once the collection goes live.
At the end of the day, a catalog is the same small workflow repeated a few hundred times, so every bit of friction you remove gets multiplied by the number of SKUs. If you want to run your next collection this way, look at PlayPause pricing and plans, pick Agency, and push a real launch through it on the 7-day free trial.
So yeah. That's my way of saying it.
Saumyajit co-founded PlayPause after years watching review and approval quietly eat creative teams' deadlines. He writes about the workflow side of video, feedback, versioning, and getting to a clean sign-off.
Related resources
Keep reading
Bring your team into one review space
Centralize feedback, lock approvals, and deliver faster, start free today.
Sign Up for Free