An account manager needs a campaign to support one retail partner. The brand team wants demand across the market. A production partner in China is asked to create the video that serves both. Before requesting a single all-inclusive quote, decide which work belongs to each budget and which commercial outcome each team is responsible for.
ONCE’s recommendation is to fund planned, reusable product assets centrally and price retailer-specific adaptations separately. Potential sales at other retailers can inform an investment decision, but an unmeasured halo should not be used to make an otherwise weak budget appear profitable.
The new signal is about the destination of value
On October 7, 2026, Circana and Best Buy Ads announced MMM findings from consumer electronics and appliance brands. Nearly 16% of media-driven sales associated with Best Buy Ads reportedly occurred at competing retailers. The announcement does not disclose the brand sample count, study period, estimation intervals or full model specification. It does not isolate video’s contribution.
Treat that as a reason to investigate cross-retailer effects, not a multiplier for your forecast. A particular model result cannot establish a universal halo rate or guarantee the return from a commissioned film.
The practical question is who can benefit from an investment, and who has been asked to pay for it. ONCE’s two-ledger approach below is a proposed planning method, not a framework validated by that study.
Define “elsewhere” before discussing the halo
The word can hide different scopes. Instacart’s shoppable video documentation, updated August 12, 2026, describes halo sales as including related products outside a campaign, such as other pack sizes or varieties within the same brand and category. That is a different question from a purchase at a competing retailer.
For your brief, specify the country, product set, sales channels, observation period and whether the result is attributed sales or estimated incremental contribution. A purchase of another SKU, a purchase at another retailer and a purchase transferred from your direct store have different implications. None should be casually added to the same total.
Ipsos MMA’s explanation of retail media measurement discusses the tension between trade funding and brand budgets, as well as the difference between attribution and incrementality. It is a useful practitioner perspective, not proof that every third-party model is reliable. Data coverage, assumptions and validation still matter.
Keep a brand ledger and an account ledger
The brand ledger asks whether the investment makes sense across the agreed market. The account ledger asks whether the brand’s team managing Retailer A can justify its assigned costs. This is the brand’s account economics, not the retailer’s own profit statement. The two views overlap; do not add them together.
Start with contribution on the brand’s revenue basis, after relevant product costs, returns, fulfillment and variable channel charges. Consumer checkout revenue is not necessarily revenue received by a brand selling wholesale. Incremental contribution then requires an estimate of what would have happened without the investment.
- Brand net incremental contribution = contribution at A + net contribution across other channels − all new media, production and measurement costs.
- A account contribution = contribution at A − media, production and measurement costs assigned to A.
Other-channel contribution can be negative. If purchases move away from a more profitable direct channel, account for that loss. Avoid counting the same commercial outcome more than once across platforms. For an additional adaptation using an existing master, separately examine avoidable new costs; do not charge an already-incurred shoot again as though it were fresh expenditure. Keep management cost allocation consistent.
An illustrative budget can expose a hidden subsidy
The following numbers are arbitrary budget units. They are neither market prices nor ONCE fees or client results.
| Planning input | Illustrative amount |
|---|---|
| Incremental contribution at A | 82 |
| Media assigned to A | 70 |
| New production | 20: shared assets 14, A adaptation 6 |
| Measurement | 10: market-wide work 8, A reporting 2 |
| Net contribution elsewhere | Unknown |
The A account view is 82 − 70 − 6 − 2 = 4. Yet the brand spends 100 overall, leaving a contribution gap of 18 before any result elsewhere. Moving the shared shoot onto a central budget changes responsibility, not the economics of the investment.
If an appropriate analysis later estimates other-channel net contribution at 24, the total becomes 6. At 10, it remains −8. The 18-unit gap is a break-even requirement in this example, not a halo result that the analyst should be pressured to find.
Until there is a credible estimate, a conservative approval scenario can exclude halo benefits while labeling them unknown. That is a planning convention, not evidence of zero effect.
Make the funding split visible in the production brief
Consider a hypothetical overseas brand sourcing a countertop appliance from China. Several retailers sell the base unit, but one carries an accessory bundle and another uses different promotional offers. A clean product demonstration may serve them all. The contents of the box and the purchase promise may not.
Ask the production partner to quote two sets of deliverables. Shared production should identify the actual product, reusable footage, source-file handover and agreed territory, duration and channel rights. Retailer adaptation should identify the exact bundle, destination, language, format, offer, revision allowance and update owner.
Central funding is sensible when the brand has named uses for the shared footage and rights to make those uses. “We might use it someday” is a weak reason to buy a large asset library. Account funding is easier to defend when the deliverable clearly fulfills that retailer’s requirements.
Do not equate a shared asset with a brandless asset. Keep useful brand cues and accurate product evidence. Separate the details that can become wrong when the destination changes: accessories, compatibility, availability and purchase instructions. Retailer commitments still need to be honored even when the brand expects broader benefits.
For the production architecture behind this separation, see ONCE’s framework for modular full-funnel video creative. Reusable footage and finished placement versions are different deliverables; a common source does not make every destination interchangeable.
Buy learning before buying a market-wide rollout
A smaller brand does not need to commission a complex model just to apply this discipline. Begin with a bounded market, product group and campaign period. Ask which channels can actually be observed, which remain missing, and which validation design the available data can support.
Where feasible, plan a geographic or store comparison in advance. Otherwise, assess whether the historical data can support MMM. A before-and-after sales increase alone cannot establish a cross-retailer effect. Require the measurement partner to explain treatment of concurrent pricing, promotions, distribution and stock availability.
For production, commission the shared shots that have real planned uses and one retailer version. Price subsequent adaptations as options. Avoid paying upfront for versions without deployment plans, and do not ask a film producer to guarantee sales outcomes controlled by media, inventory and channel execution.
For ONCE’s long-term service development, the useful assets are reusable product understanding, deployable footage and clear records of who uses each deliverable. That is a strategic judgment, not a claim of achieved margin growth or valuation gains.
Before approving the next shoot, put three numbers on the same page: all new costs, the contribution supported at the named retail account, and the remaining unvalidated gap. Those numbers can turn a vague dispute about “brand value” into a specific decision about ownership, scope and the next test.
Source notes
The Circana announcement is dated October 7, 2026; the Instacart page shows an August 12, 2026 update. The Ipsos MMA page has no clear publication date and was checked on October 14, 2026. All numerical budgeting examples and procurement recommendations are ONCE’s illustrative analysis.
