Consumer testing with AI Video
Leading global brands are now using AI video to test advertising campaign ideas, allowing them to understand customer opinion before investing heavily in production.
Ranging from still-image storyboards, to fully fledged test AI-video, specialist studios such as Myth Labs can create photorealistic creatives that mimic real advertising campaigns.This gives brands the ability to fine tune every aspect of the ad creative, and means commercials can be heavily optimised to their specific local audiences.
By equipping researchers and focus groups with video and imagery that feels realistic, brands can avoid embarrassing cultural missteps and go to broadcast with communications that are already known to be effective.
How did we get here?
Traditional Animatics
Testing creative material on consumers is nothing new. Brands often utilise classic illustrated storyboards or animatics - a hybrid between a slideshow and a basic animated video. However, the limitations of this method are numerous. Hand-drawing an illustrated animatic is notoriously slow, changes can be difficult, and the production can become quite expensive.
Furthermore, the ultimate output are sketched illustrations. Pretty, but limiting in the information consumers gain from the creative material. Illustrative campaign testing is still liable to things getting lost or missed in translation.
The development of Gen-AI
In roughly two years, AI image generation technology has matured. The days of amusing 128 pixel blurry squares are long gone. The standard for AI image generators is high fidelity and photorealistic, albeit with the occasional extra finger or toe. In recent months, the proliferation of such tools and increasing competition has led to an explosion of features that extra controls.
The new generation of tools allow artists to generatively edit out any errors, such as the aforementioned over-zealous hands that AI famously likes to produce. This added functionality has enabled AI to be used professionally, as artists can have a much greater level of control. Knowing how to ‘whisper’ the AI into doing your bidding is now a skill. In other words, creating imagery using AI for commercial use is now a process, not a game of roulette.
Localisation
The ease with which artists can now use AI to edit already AI generated images means brands can now ‘localise’ a campaign idea to get region-specific opinions on their creative.
Brands and researchers can now go to focus groups with an ad creative that has had the ‘actors’ and backgrounds swapped by the AI video production specialist. This minimises any research biases which can easily derail testing. For example, markets in south-east Asia are known to react differently to adverts that appear to feature actors from a completely different part of the world.
Cultural sensitivities regarding clothing can also be taken into consideration, which is particularly useful for testing in regions such as Saudi Arabia. Ultimately, without the ability to get feedback from a local market, brands are always leaving themselves open to missteps due to the nuances of regional sensitivities.
The Future of AI-Production for Market Testing
At the time of writing in early 2025, AI-video has made impressive progress in recent months. Nevertheless, there are still some improvements to be made, particularly in character acting.
Conveying subtle realistic emotion in anything computer generated remains notoriously difficult. The ‘uncanny valley’ effect, is a term used to describe the ‘zone’ in which CGI humans feel not-quite-right, with totally stylised (think Disney animation) on one end, and flawless realism on the other. AI-video and animation with dialogue often fall into this region, whilst AI-imagery can be made to feel completely real. Depending on the research methodology, testing animatics or board-o-matics based on still images is usually adequate.
Any shortfalls in quality will certainly be improved in the next 6 or even 3 months from time of writing. The development of more advanced algorithms through stiff competition between US and Chinese firms is accelerating the race between AI-video platforms.
Ultimately, the ability of brands to test concepts in a fast, cost-effective way allows more creative risks to be taken in early stages, without the risk of a damaging backlash.
The potential for damaging brand-image missteps in advertising are common and wide ranging. Tone-deaf cultural faux-pas, falling flat with a piece of brand activism, or indeed, accidentally crushing your customer’s dreams with a hydraulic press.
Using AI, marketers for large multi-national brands can now mitigate risks and maximise rewards from their communications. By partnering with an AI production specialist such as Myth Labs, marketers can gain a new level of control and a revolutionary new approach to understanding customer opinion.
