When you can skip a designer entirely
Not every campaign video requires a graphic designer. A short video ad for Stories, a product clip for the feed, a quick A/B test version: all of these can be produced with a video model and an internal rendering engine, without opening a project in heavy editing software. This saves days of work, and above all saves the endless loop of “one more version, one small tweak.”
The condition is a clear brief, with no requirement for complex brand identity like an animated logo, precise typography, or custom animation. In most performance ads on Facebook and TikTok, this is the case, and that is where the process saves the most.
This isn’t a shortcut that compromises quality. Every type of content has its appropriate production level, and a performance ad doesn’t need the polish of a flagship brand video. Many clients discover with relief that the initial test version, produced without a designer, is already good enough to run with a real media budget.
How it works for us, step by step
We write a short brief: product, target audience, key message. We generate a base scene using Veo 3.1 or Kling v3 as needed, and overlay text and numbers rendered separately on top of the video. The text layer is not generated by an AI model, ensuring it remains legible and precise in every version.
Then we add voiceover or background music via ElevenLabs or Stable Audio 3, and export a version ready for upload. From brief to finished file, it takes hours, not days. Before exporting, we review the text layer against the original brief, because an incorrect price or date in an ad costs more than any design round saved.
You can also stop halfway and pivot. If the first scene doesn’t work, we run another version of the same brief instead of starting from scratch. In a production with a film crew, every change of direction costs a day and budget; here, it only costs another rendering round.
For a typical performance campaign client, it looks like this: Sunday the brief arrives, Monday there are three scene versions to choose from, Tuesday the finished video is live on Facebook. With a schedule like this, there is no room for a lengthy design approval cycle, so skipping the designer stage is sometimes the prerequisite for meeting the schedule at all.
You can also run three different brief versions in parallel, not just three scene versions of the same brief. Within a day or two, you see which messaging angle resonates with the audience, even before investing an entire production in a single direction.
Where a video model alone falls short
This approach has its limits. A video model doesn’t keep a logo identical across every frame, doesn’t render accurate typography within the scene, and sometimes “interprets” the brief differently than you intended. When trying to make it generate the logo within the scene, the result is almost always distorted.
That’s why our rule is strict: the model generates the scene, not the text and not the logo. Every brand element is overlaid on a separate layer after generation, ensuring it remains identical across every produced version, even when delivering dozens of variations in the same week.
Another limitation worth noting: the model doesn’t always “understand” regulatory constraints or precise legal phrasing that must appear in the ad, such as promotional terms in fine print. This too is left to the separate text layer, rather than relying on the visual scene to handle it properly.
When you should bring in a designer
There are cases where a designer is still required: a launch campaign with a new visual language, a long-form brand video requiring consistency across multiple scenes, or custom animation that the model cannot generate. Even there, the video model doesn’t leave the picture; it becomes a tool for the designer rather than a replacement.
Our recommendation to clients: start without a designer on the test version, and bring one in only after seeing which version works. This saves design budget on concepts that would never go live anyway, reserving the designer’s time for the version that has already proven itself in the market and warrants further investment.
A common middle ground: a client starts without a designer and asks for the winning version to receive a light polish before committing a serious media budget. This works well because the designer isn’t starting from a blank slate, but rather refining something that already worked in the market, saving them time as well.
What our team does throughout all of this
The production engine runs as a unified process: brief, scene, text overlay, audio, export. There is no stage where an AI model writes on-screen text on its own, and that is the difference between a professional-looking video and one that looks like an intermediate experiment.
The team is involved at every stage: reviewing the brief, choosing the right model, and approving the result before it goes out. The tool replaces manual design labor, not human judgment, which is why every video undergoes visual inspection before reaching the client.
This is the difference between “automation” in the sense that frightens clients, and smart tools that streamline routine work while leaving people to decide what to show, to whom, and when.
Want to check if your campaign is suitable for this process without a designer? Send us a brief and we’ll provide an honest answer, including cases where you would be better off investing in a designer’s time and budget.