What generated video is really for, where it fails, and why the transparency question matters here.
Where generated video works
Five real and risk-free uses.
Generic cutaways: sky, textures, abstract movements, which are always missing at the edit.
Neutral settings and backgrounds, when the product itself is still filmed.
On-screen text variants and graphic wrappers, produced in series to test hooks.
Subtitles and language versions, which save real time.
And internal mock-ups, to validate an idea before committing to a shoot.
Where it fails
Six cases where it replaces nothing.
The demonstration of use, dealt with above.
Scale and dimensions, which the audience checks with a hand or a known object.
The product's real sound, which carries half the credibility.
Proof of use over time: the object after six months cannot be invented.
Regulated categories, where any image of the product commits the brand.
And identification: this category's audience very quickly detects a face that does not exist, and the effect obtained is the opposite of the one sought.
Transparency is becoming the norm
Four points to handle before publishing.
Synthetic content is signposted, and that obligation is evolving fast: the exact wording and its placement are checked with a lawyer, for each target market.
A voice or a face generated from a real person supposes their written authorisation, distinct from that of the original shoot.
A creator whose image is used to produce generated variants must have accepted that explicitly, with a duration and a price.
And the advertising disclosure remains due, in addition to any indication about the nature of the content.
The practical middle ground
A split that works in most cases.
The product, the hands, the use and the sound are shot.
Settings, backgrounds, graphics and text variants can be generated.
Language versions are translated, but the main voice stays human and native.
And every delivered file states what was generated, information kept with the rights.
What each approach can do
| Need | A generated video | A real shoot |
|---|---|---|
| A setting impossible to obtain | Well suited | Costly or unfeasible |
| Showing a physical product in hand | Unreliable | The very point of the format |
| A credible testimonial | Risky | Its natural ground |
What must be told to the audience
A generated image or voice is disclosed. The rule is simple: if a viewer could believe a real person is speaking, they must be told otherwise.
What remains to be verified
Disclosure rules for synthetic content and advertising are moving fast. The exact framework is confirmed with the competent authority and a lawyer before release.
The Mexican audience's angle
Four local specificities.
The audience compares a lot and checks in the comments: a visual inconsistency is pointed out within minutes.
Trust is built on the person speaking, which makes a real face more effective here than elsewhere.
Campaigns aimed at a Spanish-speaking audience across the border need a language variant, often better served by a second shoot than by a generated voice.
And the mention of synthetic content is not understood as a technical detail: it is read as a signal about the brand's seriousness.
One last marker: generated video saves time on what surrounds the product, never on the product itself. The day it is used to show the object, it does not replace a shoot, it replaces a proof, and that is exactly what the buyer came for.
