Choose your presenter video's frame from the job it must do
A wide presenter video can leave room for a product demonstration. The right format follows the placement and message, with consent and disclosure handled before release.

A presenter explaining software may need room beside them for the interface. A short mobile ad may need the product close enough to remain legible on a phone. Those requirements can lead to different frames for the same campaign. The calendar year does not decide which one will work.
Claims that horizontal video has returned or vertical video has finished its run confuse a production choice with a market forecast. YouTube's aspect-ratio guidance still identifies 16:9 as the standard for computer viewing, while its player adapts to other ratios. That supports choosing a wide frame for a specific destination. It does not predict the relative performance of every video format.
A synthetic presenter changes how footage can be produced. It leaves the framing decision intact. Before capturing an avatar or paying for generation, specify where the finished video will appear and what the viewer needs to see at the moment the presenter speaks.
Start with a frame that can explain the product
Imagine a video explaining a reporting tool. The presenter introduces a problem, then the viewer needs to compare two columns in the product. If a vertical crop makes those columns unreadable, the production needs another view of the interface. Keeping the same wide screenshot and shrinking it to fit will preserve the pixels while losing the explanation.
A wide composition can place the presenter beside the relevant area of the product. That arrangement works only if the interface remains legible at the expected playback size. A desktop editor's full-screen preview can make tiny details look more usable than they will be on the final page.
One solution is to cut from the presenter to a focused product view. The cut gives the interface the frame when the explanation requires it. The presenter can return after the viewer has seen the action. This also reduces the amount of time the generated likeness must carry the entire visual argument.
For a vertical version, rebuild that explanation around the smaller width. Show the relevant control close enough to read, with narration that names the action occurring on screen. Treat this as an adaptation of the scene. Cropping the wide master automatically cannot decide which information the audience needs.
Capture for the actual planned shot
A custom avatar should be evaluated in the composition that will be used. A close facial shot imposes different visible requirements from a small presenter panel beside a demonstration. Capture instructions and the service's supported output should be checked against that shot before recording a long session.
There is no universal capture duration that guarantees a convincing result. A vendor may require particular footage for its own process, while another system uses a different method. Follow the current instructions for the chosen product and test a short representative script before committing the campaign to that pipeline.
Consent is a separate requirement. HeyGen's current Digital Twin guidance requires a short consent recording for each video-based Digital Twin. That check supports the vendor's creation process. A company using a presenter's likeness still needs permission covering its intended use, including who may approve future scripts.
The permission boundary becomes relevant when a production changes hands. Someone may have agreed to a specific campaign without intending to authorize every later claim in their voice. Keep the approved use with the production record so an editor cannot mistake access to an avatar for permission to publish any message.
Review the whole exported message
A convincing face cannot repair a vague explanation. The script should state a claim the product demonstration can support. If the presenter says a report updates automatically, the sequence needs to show that behavior under the conditions being advertised. A polished cut that skips the required setup may create a misleading impression.
Review the export with its captions in place. Captions can cover the very control the demonstration is meant to explain. Moving the presenter or changing a cut can resolve that collision more effectively than reducing the caption size until it is difficult to read.
Distribution rules belong to the specific destination too. YouTube's current disclosure guidance requires disclosure for realistic generated or meaningfully altered content. Apply the relevant upload setting for a realistic synthetic presenter. A label does not establish that the product claim itself is accurate, so the demonstration still needs review.
Before commissioning a full series, produce one finished explanation in its intended wide placement. Watch it at the size a visitor will see, then make the vertical adaptation separately if that destination is part of the campaign. The decision to scale can follow whether each version communicates the same supported claim clearly within its own frame.