Effective advertising is largely based on psychological expertise: what captures the attention, what arouses sympathy, and what motivates action. Startup Alison.Ai, which specializes in digital advertising using generative artificial intelligence, wants to turn that into an exact science.
Recent analysis conducted by the company examined a series of video ads on social networks to discover the parameters that increased click-through rates. Among the findings: a warm, homey setting elicited results that were 62% better than public or outdoor locations. Montage or animated characters led to a 45% increase in click-throughs, and abstract symbols yielded a 43% increase.
A reverse storyline, one that begins with the conclusion and challenges viewers to understand what happened at the beginning, resulted in a 33% increase in click-throughs. And finally, partial images and silhouettes activated curiosity, leading to a 27% increase.
"The power of artificial intelligence is also the ability to see beyond fixations and concepts," explains Asaf Yanai, co-founder and CEO of Alison.Ai. "We explain the rationale to customers on a statistical level, they test us a few times, and in the end, they believe that it works.
"The market understands that AI is not a buzz word or a trend, but is everyday reality. If you want to stay at the forefront of technology, improve metrics and performance, you must use these tools. They replace people and they’re cheaper. Something that once cost $2,000, costs only a few dollars today."
"A huge AI machine"
The market share for digital advertising is constantly increasing, but - in Israel, at least - employing AI is still only for the boldest in the industry. The more conservative majority still believes in traditional methods, but that may not last long.
Alison.Ai wants to take the next step up on a global scale. Having developed an artificial intelligence platform for analyzing and creating creative content, the company is now launching a new tool for automatically creating videos up to 30 seconds long. Within a few months, it will generate videos a full minute in length.
The videos are based on the advertiser’s data, but no less important, also on the data of its competitors. How does it work? Alison's AI tools analyze elements in the existing advertisements, such as characters, colors, sounds, text, and more, and provide insights for optimizing the creative content.
The next stage is where the company offers end-to-end advertising, from analyzing and extracting insights from data of previous campaigns and from competitors, through formulating a detailed brief that includes a storyboard, copy, script, and soundtrack to creating a complete video tailored to platforms, target audiences, geographies, consumer preferences and more.
"We’re revolutionizing the online market -- in marketing, data and analytics -- using a huge AI machine," Yanai says. "Until recently, our main product was a smart brief based on a great deal of information from the client and the competition, and our clients told us it saved them 50% of their time. Now we can shorten production time by 90%. We offer a full video, which can be fed immediately into an ad campaign. This is an automated advertising agency on steroids."
No need for 30 designers
Alison.Ai was founded in 2021 by Yanai and CTO Koby Berkovich . In 2022, the company closed a pre-seed funding round of approximately $5 million. Its clients include Google, Facebook, TikTok, Zynga, LG, Warner Brothers, Ogilvy, and more.
The platform is based on customers giving the company access to their social network data. The creative elements created thus far, along with elements that have worked for the competition, are fed into generative artificial intelligence engines. The results obtained lead to a new video. "There’s no issue with copyright," Yanai says, "because we don't copy the competitor's visuals, but only use features that worked well for them."
If you’re relying on what’s already been done, how do you come up with new ideas?
"To my mind, this increases creativity exponentially. Today, it comes from brainstorming and Internet research, looking at competitors, and arriving at new concepts. We do it much faster than a human can, with no limitations."
When we spoke a year ago, you said you had no pretensions of replacing humans. That’s no longer the case.
"Initially, that wasn’t our intention, but today, the technology makes it possible. Alison users won’t need big teams. Instead of 30 designers, three will be enough. We’re improving work for the creatives. Instead of one video a week, you can produce 100 a day, based on more relevant and coherent information, and with better performance.
"We’re not becoming an ad agency, but we do work directly with clients, so they’ll have less need for digital agencies. Assisted by Alison, some will replace or reduce their activity with them, and others will leverage it to expand their activity."
Trailers for Netflix
Generative AI’s foray into the digital advertising sector has greatly lowered the barriers to entry for video creation. Without any prior experience, one can log onto a platform, write a few prompts, and generate a video. "Anyone will be able to become a video producer, and the result will be that the quantity of video will increase significantly," says Yanai.
"My expectation is that, in two years, we will see an inflation of AI-generated content. It will be a tsunami, and this is dangerous for two reasons: firstly, we users will be overwhelmed, and the user experience will be significantly damaged. And secondly, if an advertiser produces 300 videos instead of three, it will drain their budget, and inject a lot of noise into the system. Implementing analytics to separate the wheat from the chaff will be needed, and that’s where our strength, as Alison, will be."
What’s your next step?
"We will apply our capabilities to design software like Photoshop, but mainly we will work with smart televisions and streaming content. We’re in discussions with Netflix about trailers.
"Every television series has a number of trailers, sometimes dozens, the purpose of which is to attract the viewer. Why do they produce dozens? Because they don't know what will work. They do live-testing of what works, and that costs time and money. Once we’re in the picture, we’ll be able to recommend what should be included in each trailer.
"A decade ago, the concept was that one video would fit all, but today we know that doesn't work. We allow the client to enter details about their audiences -- countries, languages, age, male-female, and other features -- and the creative output is based on this. Every audience characteristic -- for example women in Alabama who use a particular product -- will have its own video."
Published by Globes, Israel business news - en.globes.co.il - on February 28, 2024.
© Copyright of Globes Publisher Itonut (1983) Ltd., 2024.