Cross-Border Product Video Market Research 06/10 · ONCE Cross-Border Product Video Market Research
Suppose a mid-priced portable power station is preparing to enter South Africa. A creative brief might ask for a more “local” setting, a more diverse cast and perhaps a local-language voiceover. Those changes may help a viewer decide more quickly whether the film is about a life they recognise. They still do not prove whether the product can solve the viewer’s problem.
This article narrows the scene to urban consumers in Gauteng, especially apartment residents, home-office workers and small workspaces in Johannesburg. The product is a portable power station with a specific model, capacity, continuous output rating, ports, charging method and service terms. This is a hypothetical decision scenario, not a client case and not a description of all South African consumers. The practical question is: how can a cross-border product film first help a target viewer recognise the use case, then turn performance claims into evidence that can be judged, and finally explain the real delivery, return and service conditions around the transaction?
The central judgement is straightforward: people and settings establish relevance; continuous product demonstration establishes verifiability; real fulfilment and service capability deliver the promise. The three layers cannot substitute for one another. Changing skin tones, décor or voiceover does not automatically create trust. Filming specifications without showing what they mean in use may not create understanding either.
“South African style” is not a usable creative setting
The South African government currently lists twelve official languages, including South African Sign Language. Statistics South Africa’s Census 2022 Provincial Profile: Gauteng shows that household language in the province cannot be reduced to an “English market”: isiZulu accounted for 23.1% in 2022, English 9.2% and Afrikaans 7.7%, with the remainder distributed across several languages.
Those figures describe household-language distribution. They do not tell us whether an individual understands English, which language they use when buying technology, what advertising style they like, or whether a particular language version will lift sales. Someone who routinely uses English at work may prefer isiZulu in a family setting. Someone from an isiZulu-speaking household may be more comfortable comparing technology in English. Age, education, urban lifestyle, product involvement, price level and viewing context can all change what a language choice means.
A useful brief therefore does not ask what “South Africans like.” It defines a narrower, testable audience and task, for example:
- adults in urban Gauteng who regularly work from home;
- people comparing mid-priced portable power stations rather than discovering the category for the first time;
- viewers on a commerce page or social platform who need to judge whether the unit can keep a router, laptop and lamp running;
- buyers whose main barriers are compatibility, safety, runtime interpretation and after-sales uncertainty, rather than awareness alone.
Only after the audience, task and media opportunity are defined do language, casting, pace and camera distance become answerable production decisions. Flags, landmarks, traditional clothing or a deliberately assembled collage of “diverse faces” cannot perform that work.
What one language study can—and cannot—tell us
In 2025, Mzamo Masito and James Lappeman published “Building Inclusivity into Advertising Strategy in South Africa: Adapting the Stern Model of Communication for Multicultural Markets” online in the Journal of Current Issues & Research in Advertising. The study used two television advertisements for household cleaning products. A total of 620 participants were assigned to English- or isiXhosa-dubbed versions; the sample was located in the Eastern and Western Cape, and many participants used English as a second or third language. The researchers combined survey responses with focus groups and expert interviews.
The isiXhosa-dubbed versions performed better on several advertising-response measures, and the study also indicates that literal translation, accent and weak language execution can become barriers. This supports a useful production hypothesis: language carries more than literal meaning; it changes tone, relationships and the judgement that “this is speaking to me.”
It does not prove that Johannesburg power-station buyers prefer isiXhosa, or that a local-language version will produce incremental ecommerce orders. The geography, category, medium and outcome measures differ. The correct transfer is not a national rule. It is a versioning test: within a defined audience, compare natural English, a relevant target-language version or a bilingual structure while keeping product facts, performer, shots, price, media opportunity and call to action as stable as possible.
Local language is therefore neither decoration nor magic. It may reduce comprehension effort when the actual audience, natural expression and buying task align. If the copy sounds translated, the performance sounds false or the central product question remains unanswered, localisation may simply make the lack of professionalism more visible.
From self-recognition to product belief: a three-layer model
Layer one: recognition—does this resemble my task?
People, spaces and sound first help viewers recognise the task. In an urban Gauteng home-office scene, the opening does not need an exaggerated power-cut drama. It can establish an intelligible work state quickly: a laptop is in a call, the router needs to stay online, a desk lamp is in use, and the portable power station sits where it can realistically be operated. The actor’s behaviour should answer “why is this needed now,” not merely display an identity.
Casting must serve the defined audience and scene. Skin tone, age, body type and household structure may help recognition, but they cannot carry a performance claim. A smiling actor holding the unit does not prove stable output. A convincingly “local” apartment does not prove that the plugs, cables, rated power or charging conditions are suitable for real use.
The same applies to sound. If the target segment genuinely calls for isiZulu or another language, voiceover, dialogue and subtitles can change. A short film should not be overloaded with several languages merely to signal inclusion. First decide who is watching, where they are watching, whether sound is on, and which language makes technical terms clearer. Then choose full dubbing, bilingual key phrases, subtitles or separate distribution versions.
Layer two: proof—can the viewer inspect the promise?
The most dangerous creative shortcut for a portable power station is to turn an attractive lifestyle scene into an implication that the unit can power anything. A persuasive demonstration keeps the causal chain visible:
- Identify the actual model and starting charge instead of substituting a look-alike sample.
- Show the specific connected devices and the relevant output ports and rated conditions.
- Keep the power station, cables, device status and key display information legible in continuous footage.
- If runtime is claimed, state the load, environment and measurement basis rather than converting theoretical capacity directly into real-world duration.
- If charging speed is shown, disclose the adapter, input method, start and end state, and test conditions.
- Put warnings and limitations where an ordinary viewer has time to read them, rather than hiding them in a final-frame footnote.
ONCE’s Chinese article on portable power-station buying reasons examines the relationship among runtime, charging and outdoor scenarios. The additional point here is that South African localisation must not change the underlying product fact. A market version may change who uses the unit, which question comes first and how it is explained, but different versions for the same SKU should not invent conflicting capability boundaries.
This is also why “authenticity” cannot be created by handheld footage, natural light or a UGC surface alone. Film language may lower the sense of advertising, but it does not replace evidence. The evidence chain still needs a disciplined claim review—covered in ONCE's Chinese-language guide to performance claims and on-screen qualifications: where the claim comes from, what the image actually demonstrates, under which conditions it holds, and what cannot be inferred from one demonstration.
Layer three: fulfilment—who delivers the promise after purchase?
A product film can explain fulfilment, but it cannot repair a fulfilment system that does not exist. For a cross-border buyer, delivery coverage, arrival time, return conditions, warranty owner, repair route, spare parts and support contact all affect trust. If those arrangements have not been established for South Africa, production should not conceal the gap behind a generic “worry-free service” line.
Effective action information is often unglamorous: name the real retailer or brand entity, state the applicable geography and warranty boundary, and show where the buyer can verify details. Video explains the route; the commerce and service systems must deliver it. If delivery, returns or repair are unstable, the problem belongs to operations rather than editing.
Rules constrain what the film may claim, not what “looks South African”
South African advertising oversight has more than one layer. The Advertising Regulatory Board’s Code is an industry self-regulatory framework rather than a substitute for statute. The current Section II of the Code of Advertising Practice, marked v2026-04-07, requires documentary evidence for objective claims before publication, with evidence relevant to the current market. Important on-screen information must also be clear and remain visible long enough for an ordinary viewer to read it.
The Consumer Protection Act 68 of 2008 is statutory law. Section 29 prohibits relevant supply-chain actors from marketing goods or services in a false, misleading or deceptive way regarding nature, properties, uses, conditions, price and other material aspects. Section 41 also addresses exaggeration, ambiguity and material omission when they amount to deception. In production terms, a naturalistic style is not permission to enlarge performance, hide conditions or blur responsibility.
In 2023, South Africa’s National Consumer Commission announced a recall of GIZZU 300Wh and 500Wh portable power stations because of a fire hazard while charging. This historical event does not mean that all portable power stations are unsafe, and it is not a judgement on any current product. It demonstrates why charging conditions, safety information, model identification, recall handling and after-sales routes are part of category trust. Lifestyle imagery cannot cover them.
Whether a specific SKU is subject to a particular compulsory specification or requires a Letter of Authority must be verified by the brand and importer with the relevant authority. This article does not generalise one technical route to every energy-storage product, and it is not complete legal or market-access advice.
What can be shared across markets, and what must be validated locally
With modular preproduction, multi-market delivery does not require every country to be filmed from zero. Nor does it mean replacing subtitles on one international master. The asset plan can distinguish two layers:
| Shareable product-fact assets | Expression assets requiring local validation |
|---|---|
| Actual SKU, ports, buttons, display and accessories | Language, accent, subtitle density and technical vocabulary |
| Continuous demonstration under stated load and conditions | Character relationships, use context and problem order |
| Evidence shots for model, capacity, continuous output and warnings | Benefit priority, humour and brand tone |
| Verified package contents and setup steps | Delivery, price, warranty, returns and CTA |
The shared layer must remain stable because it carries the same product facts. The local layer may be reordered because it carries comprehension order and transaction conditions. A multilingual master should therefore treat language versions, aspect ratios and deliverables as preproduction decisions rather than afterthoughts. Applied to South Africa, the objective is not to reserve space for twelve languages. It is to identify the segments that merit testing, then design the master so that evidence-backed differences can be made without breaking the proof.
A minimum test: measure comprehension before “liking” or sales
A useful minimum comparison does not place an English film and a completely different local-language creative into media and compare click-through rates. That design cannot tell whether the result came from language, performer, shots, price, audience selection or media opportunity.
A more interpretable test keeps the target segment, SKU, offer and media conditions constant as far as possible. It retains the same product demonstration and actor performance while changing only a native-reviewed language structure or argument order. A team might compare natural English with a relevant target-language-first version, or compare “work interruption first” against “supported-device range first.”
The first metric should be comprehension: can viewers identify which devices are supported, the conditions under which the claim holds, and where they can verify warranty and delivery? The next layer can examine claim credibility, brand attribution and stated action intent. Clicks may be observed, but they do not automatically equal incremental purchases. Sales lift requires a suitable randomised or holdout design and control over price, audience, stock, brand baseline and exposure.
If a target-language version improves the feeling that “this was made for me” but does not improve understanding of load, conditions or service, the team should not keep moving the budget into surface localisation. If comprehension improves but clicks do not, the offer, channel or fulfilment may be the next constraint—not a faster edit. The value of a falsifiable test is that it permits the team to change its mind.
Conclusion: define whose task it is before choosing how to film “South Africa”
The most common mistake in a South African product film is to let a country label substitute for a real buying situation. For an urban Gauteng home-office buyer, a credible portable power-station film must do at least three things: help the intended viewer recognise the task, provide inspectable product evidence, and explain the real transaction and service route.
People, language and setting create relevance. Continuous demonstration, conditions and source documentation create proof. Delivery, warranty and after-sales systems create fulfilment. Only when the three layers align does “local relevance” become more than art direction.
Brands preparing one product for several markets can use ONCE’s cross-border product video solution to map shared evidence shots, language versions, casting and scene requirements, and the final delivery matrix. The first decision is not a “South African style.” It is the audience’s task, the question that the camera must answer, and the promises that only a real business system can fulfil.
Research scope: This article uses materials accessible on 6 October 2026: Statistics South Africa’s Gauteng Census 2022 profile, the South African government’s official-language page, the current ARB Section II, the Consumer Protection Act, an NCC recall notice, and the 2025 Masito–Lappeman study. The language study is not a Gauteng ecommerce experiment for portable power stations, and the recall does not represent the whole category. The article offers production and testing judgements, not a preference ranking for all South Africans, a claim of client performance, or complete legal clearance.
