Thailand and Vietnam: Make the Joke Explain the Product

Cross-Border Product Video Market Research 05/10 · ONCE Cross-Border Product Video Market Research

Imagine reviewing a short film for a leave-in hair serum. For six seconds, a character rushes into frame with dramatically unruly hair. The performance, sound design and editing are memorable. The product appears in second seven, but the hand hides the pump and the dose is impossible to see. By second ten, the hair is suddenly smooth. The film never explains whether the serum goes on wet or dry hair, or whether the image represents styling, easier combing or a lasting product claim. Thai subtitles are added for Thailand, Vietnamese subtitles for Vietnam, and the two exports are called localisation.

The problem is not the humour. The problem is that the joke and the reason to buy never meet. Viewers may remember the chaotic morning without learning how to use the product or what makes the promise credible.

Our central argument is this: entertainment serves the sale only when it depends on the same product mechanism the buyer needs to understand. If viewers still cannot identify the dose, steps, use conditions, result and necessary qualification once the punchline is removed, entertainment is borrowing attention from the product. Thailand and Vietnam may share truthful footage of texture, dosing and application. The comic setup, relationships, verbal rhythm, creator identity, claim wording, notices and action information should be designed and checked separately.

Scope: one hair serum, two specific buying situations

To avoid reducing either country to a personality label, we narrow the task to a leave-in anti-frizz hair serum. The intended buyers are adults encountering the product for the first time through short-form video or an ecommerce page: a Thai-language urban audience in Bangkok and a Vietnamese-language urban audience in Ho Chi Minh City. The actual formula, pack, claim support and selling entity would still have to come from the brand's current market files.

The research cut-off is 6 October 2026. We reviewed Thai FDA material on cosmetic advertising, its updated 2024 guidance and its 2026 guide for hair and scalp product advertising; Vietnam's amended Advertising Law and implementing decree now in force; a meta-analysis of advertising humour; and several studies of Thai TikTok, Bangkok social advertising and Vietnamese creator storytelling. We did not find a field randomised experiment that holds the product, media opportunity and creative quality constant while comparing incremental purchases in Thailand and Vietnam. The conclusions below are therefore source-constrained production judgements and test proposals, not a ranking of national preferences or complete legal clearance.

A popular expression is not necessarily an explanatory one

Research on advertising humour is often compressed into the claim that funny ads sell. Martin Eisend's meta-analysis in the Journal of the Academy of Marketing Science combined 369 correlations. Humour generally improved attitude toward the ad, attention and positive affect, and the analysis reported improvements in brand attitude and purchase intention. It found no evidence for positive or negative cognitions or liking of the advertiser, however, and source credibility declined. Effects also weakened for higher-order outcomes, while moderators and research bias mattered. This is an average across heterogeneous studies, not evidence that a Thai or Vietnamese haircare film must be funny.

Regional evidence calls for an even narrower interpretation. Rawangngam and colleagues surveyed 400 Thai TikTok users in 2025 and analysed the responses with structural equation modelling. Interactivity, informativeness and customer-brand engagement were associated with perceived value, while the path from entertainment to perceived value was not significant. The model also linked perceived value, trust and purchase intention. This was self-reported observational evidence, not an experiment that manipulated humour in a film. It nevertheless challenges the easy assumption that entertainment automatically creates value in Thailand. The University of Strathclyde record links the final article and DOI 10.3390/jtaer20040319.

A separate exploratory questionnaire study of 116 Bangkok residents aged 21 to 40 who shopped through Facebook found positive correlations between perceived humorous advertising, purchase decisions and electronic word of mouth. The two studies are not true contradictions: they use different platforms, samples, variables and models, and neither proves that humour causes purchase growth. The honest production lesson is to treat entertainment as a variable to test, not as a cultural shortcut.

Direct Vietnamese video evidence is also limited. Liem Bui Quoc's 2025 qualitative case study identified humour, product review, dance-led storytelling and product-related life experience among the creator formats it observed on TikTok, and discussed the use of micro and nano influencers. It describes existing practices; it does not isolate which practice causes incremental sales. The publisher page provides the full paper under DOI 10.11648/j.ss.20251401.13. A common format is a creative candidate, not causal evidence of a national preference.

Let the joke and the product mechanism share one action

For an anti-frizz serum, useful entertainment is not a gag placed beside the product. It makes the dramatic problem arise from a genuine use situation. A character may be late while a comb catches or the ends look frizzy. The turn should not be a magical wipe that instantly changes the entire head of hair. It should reveal one pump, the product spread across the palms, the approved contact area and sequence, and the result observed under comparable lighting and combing conditions.

This gives each part of the film a distinct responsibility:

• The opening problem helps the intended viewer recognise the moment of use without inventing damage or a treatment need.
• The performance reveals the dose, contact area and order of application. A natural creator performance is still a poor demonstration if the hand blocks the product.
• The result shot keeps the hair section, brushing, lighting, moisture state and camera angle as comparable as practical. Styling, relighting and retouching must not impersonate a product effect.
• Voice and captions keep the promise within the support the brand actually has, while giving the product name, use, necessary notice and commercial relationship enough readable time.
• The call to action points to a real page, price, delivery and service path. The film cannot perform those business obligations.

The Thai FDA has publicly analysed an online serum demonstration in which eye wrinkles appeared to disappear immediately. The person was squinting before application and stopped squinting afterwards. That example concerned a facial serum, not a hair product, but the visual risk transfers: the performer's action can manufacture the before-and-after. With hair, a change in brushing, key light, selected section or styling can create the same false impression.

Thailand: the film can perform, but the performance cannot enlarge the claim

As checked on 6 October 2026, the Thai FDA stated on 8 September 2025 that cosmetic advertising does not itself require an advertising licence, but must be fair, truthful and non-misleading. False, exaggerated and therapeutic or disease-curing claims are not allowed, and the product should still be properly notified. The regulator's cosmetic advertising area lists updated 2024 guidance, and on 25 May 2026 it published a specific guide for advertising hair and scalp cleansing and care products. The production implication is not to add a visual stamp that suggests general FDA approval. It is to keep the product identity, notification status and supported claim aligned.

A Thai version may dramatise getting ready in humid weather, but the approved brand material must determine what “anti-frizz” means: an immediate grooming appearance, styling hold, or a metric under specified test conditions. If the brand has only texture and use documentation, the film should remain specific: number of pumps, wet or dry hair, mid-lengths or ends, and whether blow-drying follows. If an instrumental or consumer study exists, the team should confirm the tested formula, dose, time point and permitted wording. A light flare and a surprised face cannot substantiate “all-day repair”.

The comic tone should follow the audience's task. A creator-led film for people who already understand the category can move rapidly from the problem into the action. First-time buyers may need the product category and application area earlier, so they do not mistake a leave-in serum for a rinse-out conditioner or scalp treatment. When a client asks for something “more Thai”, the first variables to change are linguistic naturalness, the situation and performance rhythm—not an untested increase in exaggeration.

Vietnam: the creator must first be an identifiable advertising source

Vietnam's Law No. 75/2025/QH15 took effect on 1 January 2026. It brings people who directly advertise, recommend or endorse products online within the definition of persons conveying advertising. Influencers must verify the advertiser's reliability and inspect relevant product documentation; they must not introduce a product they have not used or do not understand. They must also announce the advertising immediately before and during the activity. Online and social advertising should be clearly distinguishable from non-advertising content. A film made to look like a casual recommendation does not remove the commercial relationship, and UGC styling is not the same thing as organic user speech.

For the creative, disclosure should not flash only at the end or disappear behind a joke, sticker or dialogue. A creator can begin with a recognisable problem, but the advertising identity, product name and task should appear before viewers build a false understanding. The brand should give the creator verifiable product documentation and distinguish real use experience, approved brand facts and scripted performance in the collaboration records. A conversational delivery must not fabricate a long-term personal experience.

Vietnam's Decree No. 342/2025/NĐ-CP took effect on 15 February 2026. Its cosmetic advertising provisions require the cosmetic name, function or use, the name and address of the organisation or person responsible for the product declaration, and warnings required under applicable international agreements. The advertising must not make a cosmetic look like a medicine and must remain consistent with the product's nature, classification and declared function. The exception allowing a sub-30-second television advertisement not to voice the warning, provided it remains in text and the information is supplied fully, is specific to television. It should not be generalised to every social video.

Vietnamese caption design therefore starts before translation is squeezed into the edit. The product name, declaring entity, function and necessary notice need a readable hierarchy in the vertical safe area. However fast the creator speaks, mandatory and decision-critical information cannot be accelerated with the punchline. The denser the entertainment, the more important it is to reserve protected information time in the script.

How to rebuild the same assets for Thailand and Vietnam

This is not one “international” film exported with two caption files. The shoot preserves reusable product evidence; post-production rebuilds the joke, argument and disclosure for each language. ONCE's ecommerce product video production approach plans listing films, detail demonstrations, ad cutdowns and language versions together. Its AIGC commercial production page features a Herbal Essences botanical haircare project and explains why shots that establish products and precise actions should use live action or approved assets. The page is useful for observing the division between visual direction and product accuracy, but it provides no Thai or Vietnamese market-effect evidence.

When a project also involves religious context, certification or different packs, see the previous essay, Halal Is Not a Visual Style: Product Video for Indonesia and Malaysia. That article addresses credentials and cultural symbols; this one addresses entertainment, demonstration and advertising identity. Both point to the same principle: a local cue cannot substitute for a product fact.

A minimum test: measure comprehension before liking

Proposed experiment; not yet conducted: create two 20-to-25-second versions separately for Thailand and Vietnam. Do not pool the countries into one result. Keep the product, price, landing page, person, main shots, placement and opportunity to see as comparable as possible. Change only the order of entertainment and demonstration.

• Version A uses a three-to-four-second situational joke to establish the problem, then shows the complete dose and application, followed by the same claim, disclosure and call to action.
• Version B establishes the product category and one clear demonstration first, then lets the joke become the narrative reward after use, followed by the same claim, disclosure and call to action.

Do not choose a winner from completion rate alone. First ask intended buyers: What is the product? How and where is it used? What does the image demonstrate, and what does it not demonstrate? Who is making a commercial recommendation? Who provides the service after purchase? Failure on these questions means the film has not completed its basic communication task. Once comprehension is acceptable, compare credibility, brand attribution, click or choice. Claims about incremental purchase require suitable randomisation, a holdout or another design capable of identifying incrementality.

What would change our recommendation? If Version A preserves product comprehension, disclosure recognition and credibility under comparable exposure while repeatedly improving downstream choice, the comic opening deserves to stay or expand. If it improves only viewing and liking while more people misunderstand the use, effect or advertising identity, shorten the entertainment or move the joke inside the demonstration. If Version B is clear but loses too many viewers in the first seconds, do not bolt on an unrelated gag; make the product problem itself more dramatic.

Conclusion: localisation redesigns understanding, not just the punchline

Entertainment-led, narrative and creator-led product content exists in both Thailand and Vietnam. Common use is not causal effectiveness, and it does not mean the same performance works in both markets. For an anti-frizz serum, the most reusable assets are the actual sample, dose, texture, application and comparable result. The variables most likely to need separate builds are language, relationships, comic timing, creator responsibility, claims, notices and purchase conditions.

Before filming, the brand should prepare six inputs: the actual market SKU and pack for each country; approved claims and qualifications; notification or declaration files; the complete use process; creator collaboration and disclosure rules; and the landing page, delivery and customer-service path for each market. ONCE can use these inputs to separate the shared demonstration, Thai edit, Vietnamese edit and essential reshoots before deciding how loud the entertainment should be.

Discuss Thai and Vietnamese product video versions with ONCE: define the product, audience and purchase question before choosing the story, creator and joke.

Sources and access scope

Official sources: the Thai FDA statement of 8 September 2025, its updated 2024 cosmetic advertising guidance area, its hair and scalp advertising guide of 25 May 2026, and its public analysis of a manipulated-looking online serum demonstration; Vietnam's Law No. 75/2025/QH15 and Decree No. 342/2025/NĐ-CP. Research sources: Eisend's meta-analysis of advertising humour; Rawangngam and colleagues' Thai TikTok survey and structural equation model; Chinda and Anantachart's exploratory Bangkok questionnaire; and Liem Bui Quoc's qualitative case study of Vietnamese TikTok creators. We relied only on the text, full articles or abstracts actually available. Correlations and survey models are not presented as randomised causal evidence, and observations of common formats are not presented as proof of commercial lift.