How Should a Multi-SKU Brand Split Its Photo and Video Budget?

Video Production Budget Decisions 03/10 · ONCE Guide to Video Production Budget Decisions

When 12 SKUs enter one production brief, the easiest move is to divide the budget by 12. Every product gets the same number of images, the same video duration, and the same amount of studio time. The spreadsheet looks fair. The deliverables may not be: the hero products lack enough coverage to explain a multi-step function, while long-tail colour variants receive lifestyle scenes that no channel will use.

The budgeting unit should not be the file, and the calculation should not begin with a price per SKU. A better sequence is to identify the work that the entire batch must share, add variable work according to the job each hero, core, and long-tail SKU has to perform, and then test those choices against available cash, inventory, and contribution per order. That process may show that some upgrades are justified, some long-tail coverage should be produced internally, and some unready SKUs should leave the current batch.

This article uses a complete hypothetical CNY model. Every rate is an assumption for calculation, not a market quote or an ONCE quote. Replace each number with quotations that cover the same scope, rights, and acceptance criteria before making a purchase decision.

Hold the scenario constant: 12 SKUs are not 12 identical jobs

Assume a kitchen-products brand selling into the US ecommerce market plans to update 12 product pages for a six-month use period:

• Three hero small appliances need standard product images, detail images, a step-by-step function demonstration, and a simple lifestyle context.
• Six core specifications or accessories need consistent product images, close-ups of their differences, and one key-action clip.
• Three long-tail colours or replacement parts only need accurate identification, dimensions, and compatibility coverage; they do not need talent.
• Both photos and video are required, but still-image retouching, motion continuity, and lighting across an action are accepted separately.
• Completion is measured by usable product-page tasks, not by the number of exports.

The brand supplies approved product facts, a SKU list, compliance copy, production samples, spares, and a person authorised to approve function and appearance on set. The production team converts those inputs into shots, manages shared and independent production states, captures and finishes the work, names the files, and handles technical acceptance. Talent, music, special props, and additional versions count only when they appear in the written scope.

Shopify's current product media documentation treats images, video, and 3D models as product media and explains that a featured media item appears in several storefront contexts. Amazon's current US guidance asks product video to be descriptive, demonstrative, and product-focused. Those sources help define different media jobs; they do not tell a brand what it should spend. The budget still has to be built from this project's tasks and risks.

Separate shared batch cost from variable SKU cost

A batch does not share every production task. It may share a SKU audit, art direction, a set build, a master lighting test, file rules, and batch-level quality control. Product cleaning, function reset, unique angles, actions, retouching, and revisions still increase with each SKU.

Use two formulas:

Batch cost = shared cost + sum of variable SKU costs + risk reserve

Cost per completed SKU = actual batch cost / SKUs whose required tasks passed acceptance

The second formula matters. If 12 products arrive and the team exports 180 files but only 10 SKUs have complete product-page coverage, the denominator is 10—not 12 and certainly not 180.

For the standard hybrid option, assume CNY 15,200 of shared work:

The hourly rate and package assumptions are model inputs. A real quotation should state tax, studio, equipment, transport, talent rights, revision rounds, source files, travel, and the treatment of late samples. Prices for different scopes are not comparable.

Three feasible spends: all cover 12 SKUs at different depths

Lean option: CNY 39,600

Assume CNY 12,000 of shared preparation and CNY 2,300 of standardised work for each of 12 SKUs: consistent base product images, necessary difference images, and one short clip that makes the key function or specification understandable.

CNY 12,000 + 12 × CNY 2,300 = CNY 39,600

This option fits products with similar size, materials, and actions when every sample is ready, product facts are clear, and the current objective is the minimum usable set of product-page media. It excludes talent, a custom lifestyle set, complex camera movement, and a narrative designed specifically for the hero products.

It is not a discounted standard option. If three hero products require a full operating sequence, the CNY 2,300 standard task is insufficient; compressing every action does not make the required proof complete.

Standard hybrid option: CNY 64,900

The standard option assigns variable cost by task:

The average is CNY 5,408 per SKU, but it is not a price for any individual SKU. Hero and long-tail products buy different jobs. The average is useful only for assessing cash efficiency across the batch.

Enhanced option: CNY 103,600

Assume the enhanced option adds two lifestyle sets, talent with a defined usage scope, more complex movement or visual treatment for the three hero products, and more channel versions for core SKUs:

Shared work and rights CNY 24,000 + 3 hero × CNY 14,000 + 6 core × CNY 5,000 + 3 long-tail × CNY 1,800 + reserve CNY 2,200 = CNY 103,600

This is justified only when each addition performs a defined job: talent must demonstrate scale or interaction, a second set must cover a genuinely different environment, or a confirmed channel must require a separate format. If the rationale is merely to look more premium without a distribution plan, rights term, or different acceptance test, there is no adequate procurement case for the upgrade.

Adobe's March 2026 guide to ecommerce product photography distinguishes individual, detail, scale, lifestyle, and collection images. The taxonomy helps define jobs; vendor material is not evidence of budget effectiveness. Generative tools may be evaluated for non-evidentiary background changes or layout derivatives, but product structure, compatibility, and successful function still require acceptance against the real product. Generation volume is not completed-SKU volume.

Why separate projects may not cost less

Now hold the standard tasks constant but commission 12 separate projects. Assume each one repeats CNY 2,625 of briefing, setup, testing, and file administration:

12 × CNY 2,625 + the same CNY 43,800 of variable SKU work + CNY 6,000 reserve = CNY 81,300

The difference from the CNY 64,900 batch is CNY 16,400. It is not a market-wide batch discount; it is the modelled cost of repeated preparation. If samples mature at different times, or the products require unrelated materials and sets, multiple production blocks may reduce waiting and rework. In that case, forcing unready products into one day simply to preserve shared cost is a false economy.

An equal CNY 60,000 allocation is a useful counterexample: CNY 5,000 per SKU may overfund talent coverage for long-tail colours while remaining insufficient for a hero product's multi-step operation, lifestyle context, and explanation in post. A reasonable total does not guarantee a reasonable allocation.

Calculate the production floor, then the cash ceiling

If all 12 SKUs must receive basic stills and usable video, the modelled production floor is the CNY 39,600 lean option. The cash ceiling should come from money actually available after protected needs, not from imagined future revenue.

Assume the brand can allocate CNY 90,000 to the content update but must retain:

• CNY 15,000 for channel launch and first distribution;
• CNY 8,000 for sample fixes and possible pickup preparation;
• CNY 5,000 as an operating cash buffer.

The external production cash ceiling is therefore:

CNY 90,000 − CNY 15,000 − CNY 8,000 − CNY 5,000 = CNY 62,000

The CNY 64,900 standard option exceeds that ceiling by CNY 2,900. The right response is not to demand identical scope for CNY 62,000. Change ownership instead: move the three long-tail tasks, modelled at CNY 4,200, to the client's internal team. Keep professional production for the hero and core products and retain the CNY 5,900 risk reserve. External cash spend becomes CNY 60,700.

If internal work takes two hours per SKU at an opportunity cost of CNY 180 per hour, the internal opportunity value is 3 × 2 × CNY 180 = CNY 1,080. Total economic cost is about CNY 61,780 while cash spend remains CNY 60,700. That CNY 1,080 is a cash saving only if the work actually avoids overtime, a contractor, or another payable cost. Otherwise, it is time redirected from other work.

If available cash is below CNY 39,600, the valid choices are to reduce this batch's SKU count, move approved tasks in-house, or delay unvalidated work. Keeping every task, SKU, and acceptance requirement while squeezing every line item is not a fourth option.

Upgrade thresholds for hero and core products

Contribution per order is not the retail price. Here it means net selling price minus the variable costs associated with the order: product cost, platform fees, fulfilment, expected returns, necessary incremental acquisition cost, and other relevant variable expenses. Shopify's 2023 explanation of the contribution-format income statement likewise calculates contribution after variable expenses. The brand's finance or operating owner should replace the assumptions with actual data.

Assume CNY 600 contribution per hero order, CNY 300 per core order, and CNY 120 per long-tail order.

• Moving a hero SKU from the CNY 2,300 base task to CNY 6,800 adds CNY 4,500. Across three heroes, CNY 13,500 of added production requires CNY 13,500 / CNY 600 = 22.5, rounded up to 23 incremental orders across the group, not 23 orders per SKU.
• Moving six core SKUs from CNY 2,300 to CNY 3,200 adds CNY 5,400. That requires CNY 5,400 / CNY 300 = 18 incremental orders.

These are hurdles, not sales forecasts. If the three hero products have only 15 sellable units combined during the six-month period, a 23-order hurdle is physically impossible. Unless lifestyle coverage also satisfies a confirmed channel, sales-support, or avoidable-cost requirement, reduce the upgrade.

The adjusted standard option costs CNY 21,100 more in external cash than the lean option. Sensitivity looks like this:

Keep the zero-lift case visible. If video creates no attributable incremental contribution, the extra CNY 21,100 does not recover itself. The buyer can only justify the work through another necessary deliverable; visual polish cannot be rewritten as a realised commercial result.

What happens when samples are not ready

Before the lock date, every SKU should pass six checks: sample appearance, function, accessories, packaging or labels, approved product facts, and availability of the person responsible for acceptance. The deadline depends on shipping and project conditions; it is not a universal industry rule. This example uses T−7 days.

Assume one core SKU and one long-tail SKU exit before T−7 and their variable work has not begun. Standard cost falls by CNY 3,200 + CNY 1,400 to CNY 60,300. Ten SKUs complete, so cost per completed SKU is CNY 6,030. That is higher than CNY 5,408 with all 12 complete because shared preparation does not decline in proportion to the SKU count.

If the two products fail after lock, assume for the model that each has consumed CNY 1,000 of testing and reserved time. Spend becomes CNY 62,300, or CNY 6,230 for each of ten completed SKUs. The CNY 1,000 is still an assumption, not a cancellation-rate benchmark. A real contract should state how cancellation, sample replacement, retesting, and set restoration are charged.

Write the exit rule before booking:

1. A SKU missing a production sample, approved product facts, or essential accessories at T−7 exits the current batch.
2. If only a non-functional colour or packaging detail is missing, the SKU may enter a restricted appearance task but cannot be marked complete.
3. A replacement SKU enters only when its size, material, action, and file task can use the same production state.
4. A change after lock triggers confirmation of actual incurred cost and pickup restoration cost.
5. The batch report retains separate columns for planned SKUs, SKUs admitted to production, and completed usable SKUs.

Photos and video can share preparation, not acceptance

Photography and video can share product facts, props, colour references, a sample manager, and parts of the set. They should not automatically share one lighting state, time block, or acceptance action. A still may be tuned for a single highlight and retouched frame; video requires reflection, focus, and product behaviour to hold through movement. ScreenSkills' current production coordinator responsibilities connect scheduling, resources, call sheets, and progress records. A multi-SKU commercial shoot differs in scale, but it still needs written ownership for inputs and changes.

Give each SKU separate fields for:

• Photography: angles, details, colour reference, retouched-image count, and permitted product corrections.
• Video: action, successful-function criteria, aspect ratio, sound or interface, and preferred take.
• Shared preparation: art direction, sample cleaning, set, styling, and file naming.
• Independent reset: lighting, camera, product state, consumables, cleaning, and action continuity.
• Acceptance owner: the brand approves product facts; production approves technical quality and files; the account owner checks final platform display.

“Captured” is not “complete.” Completion means each required task maps to an openable file, product facts and appearance have been approved, omissions have an explicit reason, photo and video separately meet the agreed specifications, and masters, outputs, naming, versions, and revision scope follow the contract.

A fillable budget sheet for real quotations

Put these fields into one comparison sheet:

Pre-write the stop conditions as well: the production floor exceeds the cash ceiling; hero inventory or contribution cannot cross the upgrade hurdle; essential samples and facts miss the lock date; or the enhanced option lacks a confirmed channel, rights scope, or acceptance job. When any condition is true, reduce scope or delay. Do not make craft quality absorb the project's cash risk.

For an unvalidated single-SKU product, continue with the series guide to a first product-page video budget. If the hero product is already established and the question is whether to add lifestyle scenes and talent, see the premium product video upgrade thresholds.

To scope a multi-SKU batch with ONCE, share the product list, target market, channels, existing assets, sample status, launch timing, acceptable cash budget, and non-sensitive assumptions for contribution and inventory. We can first map shared cost, hero/core/long-tail jobs, separate photo and video states, internal pickups, and exit rules into one production scope—then decide whether to batch, split, or defer SKUs. Quotation, availability, and output depend on the confirmed scope.

Source note: On October 9, 2026, this article checked Shopify's product-media and file documentation, Amazon's US product-video guidance, Adobe's March 2026 ecommerce photography guide, Adobe Firefly business materials, and ScreenSkills' production-coordination responsibilities. All CNY rates and budgets are disclosed assumptions for calculation, not market averages, supplier quotations, or an ONCE price list.