Instant Studio Playbooks

How to localise a global campaign for every market — without reshooting it

2026-07-07T00:00:00.000Z

You've shot a beautiful hero campaign. It tested well, the brand team loves it, and now it needs to run in twelve markets. That's where the maths stops working: every market wants local talent, a local setting, local copy — and the traditional answer is a reshoot, a local production partner, or a compromise. Reshoots multiply the budget. Local partners multiply the coordination. And the compromise is usually subtitles over footage that everyone can tell was made somewhere else.

There is now a fourth option: treat the campaign as a system of components and regenerate it per market, using AI production workflows anchored to your original creative.

Why translation isn't localisation

Swapping the voiceover and headline is the cheapest path, and audiences see through it instantly. What actually makes creative feel locally made:

Talent — presenters and background casting that match the market
Setting — a Jakarta street doesn't read like a Munich one; product scenes need local context
Cultural references — props, gestures, humour, holidays, even the food on the table
Format — some markets live on 9:16 short-form; others still need 16:9 and print
Copy — transcreation that carries the idea, not word-for-word translation

A localisation workflow has to touch all five. That's why it historically meant a reshoot — no editing suite could change the talent or the location after the fact. AI production can.

The workflow, step by step

Here's how a campaign localisation workflow runs inside a managed AI production system:

Anchor on the source creative. The original campaign — key visuals, product shots, brand guidelines, tone — becomes the reference the whole system works from. Brand colours, product accuracy and composition rules are constraints, not suggestions.
Generate the scene per market. The environment around the product is regenerated for each market: architecture, landscape, light, interiors. The product itself stays pixel-accurate to the source.
Layer market-specific props and details. The details that make a scene belong somewhere — signage, plants, vehicles, seasonal cues — are added per market.
Match talent to market. Presenters and background talent are generated or matched per market, with documented rights and consent handling for any real-person likeness.
Localise the copy. Headlines and scripts are transcreated, then set in the local typography — including scripts like Traditional Chinese, Thai or Arabic that break naive templates.
Human review, market by market. Local reviewers and the brand team approve each market's output before anything ships. The AI does the production; people keep the judgement.

One automotive campaign we produced this way started from a single source image of the vehicle and became localised video content across markets — different languages, settings, locations and background talent — without the vehicle ever being reshot.

What changes for the team

The practical effect is that localisation stops being a production problem and becomes a review problem. Instead of coordinating shoots, your regional teams review and approve. The centre keeps brand control; the markets keep cultural judgement. Timelines compress from months to weeks, and adding a thirteenth market is an increment, not a new project.

It also changes what you can afford to localise. When adaptation is this cheap relative to a reshoot, the long tail of assets — CRM banners, in-store screens, social cutdowns — gets localised too, instead of running in English everywhere except the hero film.

When this approach fits — and when it doesn't

It fits when you have strong source creative, clear brand guidelines, and markets that deserve more than subtitles. It's strongest for product-led categories — automotive, consumer electronics, FMCG, beauty — where the product must stay exact while everything around it adapts.

It's the wrong tool when the campaign concept itself doesn't travel (no production technique fixes a culturally wrong idea), or when a market demands creative developed from local insight rather than adapted from global. Localisation scales a good idea; it doesn't replace local strategy.

The teams getting the most from this treat it as a workflow they run every campaign through, not a one-off experiment. The second campaign through the system is faster than the first; the fifth is routine.

50 ad variants from one brief: how high-volume creative production actually works

2026-06-23T00:00:00.000Z

Every performance marketer knows the pattern: a new ad works for two or three weeks, fatigue sets in, performance decays, and the channel team asks for fresh creative. The creative team, reasonably, points out that they shipped the last batch a month ago and there are only so many hours in a week.

The demand side of this problem isn't going away — ad platforms reward creative volume and variety, and their own optimisation systems work better with more variants to allocate against. So the supply side has to change. That's what high-volume variant production is: a way to make one brief produce fifty testable ads instead of five.

Think in a matrix, not a list

The mistake most teams make is briefing variants one at a time — "can we get a version with the blue background?" A variant system starts from a matrix instead. Pick your dimensions:

Message — price, quality, social proof, urgency, problem-led, benefit-led
Format — 9:16 video, 1:1 static, carousel, 16:9, story with UI overlays
Visual direction — studio product shot, lifestyle scene, UGC-style, illustration
Audience — new customer vs. returning, segment-specific hooks
Market — language, local context, seasonal timing

Six messages × four formats × three visual directions is seventy-two variants from one brief. Nobody hand-produces seventy-two ads. A production workflow generates them; your team curates.

Why templates aren't the answer

The first generation of "scale" tools solved this with templates: lock the layout, swap the product image and headline. The output is technically varied and creatively identical — and audiences tune it out almost as fast as a single ad, because the idea never changes, only the wallpaper.

Real variants change the creative substance — the hook, the scene, the presenter, the demonstration — while holding the brand constant. That's the harder problem, and it's where AI production workflows earn their keep: the brand system (colours, type, product accuracy, tone, claims) is embedded as a constraint in the workflow, and the generation explores everything else.

The workflow

In a managed production system, a variant run looks like this:

One brief, structured. Campaign goal, product truth, approved claims, brand rules, the variant matrix you want to explore.
Generation against the matrix. The workflow produces the full grid — every message × format × direction combination — with the product kept exact and the brand system enforced.
Automated checks, then human QA. Outputs pass through brand and claim checks, then a human review pass kills the ones that are off-tone, off-brand or just not good. Expect to publish a subset of what was generated; the economics still work because generation is cheap and review is fast.
Ship, measure, regenerate. Winning directions become the seed for the next batch. Fatigued angles get retired. The matrix evolves with the data.

The unit economics are the point: when a variant costs a fraction of a hand-produced ad, testing ten hypotheses per week stops being a luxury.

What your team actually does

None of this removes people from the loop — it moves them up the stack. Strategists design the matrix. Creatives set the visual directions and judge the output. Channel managers decide what gets budget. The work that disappears is the mechanical production of the 40 variants between the first and the last — which is exactly the work nobody enjoyed.

A good test for whether your team would benefit: count how many creative hypotheses you wanted to test last quarter versus how many you actually shipped. If the gap is large, the constraint is production capacity — and that constraint is now optional.

UGC-style product video at scale: a workflow for TikTok, Reels and Douyin

2026-06-09T00:00:00.000Z

Short-form social has a production problem hiding inside a creative insight. The insight: polished brand films underperform on TikTok, Reels and Douyin, while creator-style content — handheld, direct-to-camera, demonstrative — wins attention and converts. The problem: that "authentic" content still has to be produced, and the platforms reward accounts that post constantly.

For an ecommerce team with hundreds of SKUs, the maths is grim. Creator partnerships are slow to negotiate, inconsistent in quality, and don't scale across a catalogue. In-house shooting burns out a content team fast. Most brands end up posting far less than the channel rewards, with a fraction of the catalogue ever getting video at all.

The way out is to treat short-form product video as a production workflow with product images as the input — not as a series of shoots.

What "UGC-style" means when AI produces it

UGC-style is a format, not a sourcing strategy: direct-to-camera presentation, natural settings, product demonstration, platform-native pacing and text overlays. AI production can generate this format from your product assets:

AI presenters deliver to camera in the market's language, with rights cleanly handled — no likeness ambiguity, no renegotiation when the campaign extends
Product demonstration scenes are generated from catalogue images — the product in hands, in use, in context — while staying pixel-faithful to the real item
Platform-native packaging — 9:16 framing, hook in the first second, captions and overlay text in local script — is applied in the workflow, not in a manual edit pass

The non-negotiable is product accuracy. A generated video that misrepresents colour, size or function is a returns problem wearing a marketing costume. This is why the workflow anchors on your actual product imagery and why QA reviews outputs against the product truth, not just for aesthetics.

From product image to published video

A working pipeline looks like this:

Input: the product feed. Catalogue images, product attributes, approved claims, brand tone. For most SKUs this already exists — no new assets needed.
Scene and script generation. For each product, the workflow generates demonstration scenarios and short scripts matched to the platform and market.
Video generation with an AI presenter or product-only format. Presenter-led for products that benefit from a human hand and voice; product-in-context for the rest.
Human review. A reviewer checks product accuracy, claims and tone. At steady state this is minutes per video, not hours.
Publish, read the data, iterate. Hooks and formats that hold attention get more SKUs pushed through them.

Teams running this kind of workflow have cut per-video production time from hours to minutes — which is the difference between videoing your ten hero products and videoing the whole catalogue.

Where the humans stay

Three places, permanently. Strategy: which products, which hooks, which markets — the workflow executes taste, it doesn't supply it. Review: every published video passes a human check on accuracy and brand. Rights and compliance: presenter likeness, claims substantiation and platform rules are process questions, and they need owners — especially in regulated categories.

That's also the honest boundary of the approach. If your brand's whole positioning rests on named creators and their communities, AI presenters complement that programme; they don't replace the community. What they replace is the impossible middle: the two hundred competent, on-brand demonstration videos per month that no creator programme was ever going to deliver.

The channel isn't going to want less video next year. The question is whether your production system is shaped like the demand.

One photo, ten SKUs: the unglamorous work that gets a product catalogue online

2026-05-26T00:00:00.000Z

Most conversation about AI creative production is about the glamorous end: hero films, campaign concepts, localisation across twelve markets. Then a real brief lands, and it looks like this: a client who sells industrial fasteners online is refreshing their product photography. There will be roughly 500 raw camera files. Out of those, around 3,000 finished images need to exist on the website — sized, labelled, watermarked, optimised, named and tagged. Turnaround: as soon as possible, because the site launch is waiting on the images.

Nobody puts this work in a portfolio. It decides whether the site ships on time.

The anatomy of a bulk product-image job

Briefs like this are more interesting than they look, because the multiplication is doing all the work:

One photo, many SKUs. A hex bolt comes in ten sizes. Nobody shoots ten near-identical bolts — you shoot one, and produce ten images with the size rendered on the image: M8, M10, M12. One photo becomes ten catalogue entries. This is where 500 shots become 3,000 images.
Brand furniture. Every image carries the company logo as a watermark — same position, same opacity, every time.
Platform spec. Exact pixel dimensions, exported as both JPEG and WebP, under a file-size budget (say, 100KB) without visible quality loss — because page speed is a conversion and ranking factor, and product pages live or die by it.
Findability. Filenames follow a naming convention from a spreadsheet. Alt text goes into the metadata, also from a spreadsheet — which is simultaneously an accessibility requirement and free search visibility most catalogues never claim.
RAW input. The photographer hands over CR3s, not JPEGs. Everything above sits on top of consistent RAW processing — exposure, white balance, colour — so image 2,847 matches image 12.

None of these steps is difficult. That's the trap.

Why it goes wrong when humans do it by hand

Three thousand images times seven operations is over twenty thousand operations, every one of them trivial and none of them allowed to be wrong. Manual production fails in predictable ways: the watermark drifts a few pixels between operator one and operator three; a batch gets exported at the old dimensions after the spec changes; filenames diverge from the spreadsheet somewhere in row 900 and surface as a mess at upload time; alt text gets quietly skipped when the deadline tightens, and the SEO benefit evaporates before anyone notices it was gone.

And then the spec changes — it always changes. "Actually, can we do 1200px, and move the size label to the top-left?" With hand production, that sentence costs weeks. It should cost one overnight re-run.

Treat it as a workflow, not a task list

The fix is to stop thinking of this as 3,000 editing tasks and set it up as one workflow that runs 3,000 times:

The spreadsheet is the brief. SKUs, size variants, filenames, alt text — the client's own spreadsheet becomes the machine-readable spec the workflow executes against. If a row is wrong, you fix the row and re-run, not the image.
RAW processing as a profile, applied uniformly, so consistency is a property of the system rather than of whoever was working that day.
Variant expansion from the size matrix — one master image per product, every size variant generated with the label typeset to brand rules, not stamped on by hand.
Watermark, resize, export as deterministic stages: both formats, compression tuned to the budget, dimensions from the spec.
Naming and metadata written straight from the spreadsheet — alt text embedded, filenames exact.
Human QA on samples, not on everything. Reviewers check a statistical sample per batch and anything the system flags as anomalous; people supply judgement, the workflow supplies repetition.

This is the same shape as generating fifty ad variants from one brief: inputs × a variant matrix × brand rules, executed by a system, checked by people. The subject matter is less glamorous; the economics are identical. AI earns its keep in the judgement-adjacent corners — consistent RAW-to-clean processing, background standardisation, smart cropping, drafting alt text when the spreadsheet doesn't exist yet — and disciplined automation does the rest. What the client experiences is simpler: the spec goes in, the finished catalogue comes out, and a re-run costs almost nothing.

The boring stuff compounds

A catalogue produced this way is faster (WebP under budget on every product page), more findable (complete, consistent alt text and naming), and more trustworthy (every image visibly from the same family). And when the product line grows or the photography refreshes next year, the workflow is sitting there, ready to run again — which is the real difference between buying an editing job and building a production capability.

If you're writing a brief like this, the checklist above is the brief: input format, variant logic, labelling spec, watermark, dimensions, output formats and file-size budget, naming convention, alt text source. Get those eight things on one page and you're most of the way there. Running it at volume, on deadline, without drift — that's the part we do every day.