Opinion
Can GenAI really craft effective advertising campaigns?
Summary
- The way GenAI is designed to work, it doesn’t deliver differentiated output. Nor does it have a grasp of the emotional stimuli that effective ads, like Apple’s, tend to thrive on. The technology will improve, but advertising would still need human ingenuity.
Why is the masala kept in a separate pouch inside the noodles pack? What lessons does it hold for those developing AI algorithms?" These were questions I recently asked a class of management students. Though I told them not to use ChatGPT, at least 20 of the 60 students in the class used it to answer the question. How did I find that out?
I did not use any sophisticated technological tool to identify assignments written by humans and those written using ChatGPT. Although the words used in the ChatGPT-written submissions varied slightly, they had the same arguments.
More notably, they had the same logical structure for those arguments. Much like the standardized output of an assembly line at a factory, what ChatGPT rolled out looked very similar. The answers lacked differentiation.
Generative AI (GenAI) is possibly one of the greatest leaps taken by the world of technology. It has already replaced a few human functions in many organizations. There are many who believe that one function that is ripe for GenAI to dominate is the creation of advertisements.
An ad is mostly about an image, a headline and few lines of copy. The generation of these should be easy for GenAI technology. Today, organizations that use human talent spend much money and time on creating ad campaigns.
Also read: Marketers and brands have good reason to celebrate the rise of AI
An expert photographer will charge a very high fee and take several days to shoot an ideal image. With GenAI, images can be generated in a matter of minutes, and that too at scale, resulting in huge cost savings.
It is true that GenAI is still unable to generate images that are fully accurate representations of original products. Given the pace of technological advancement, however, it is only a matter of time before GenAI perfects the accuracy of the images it creates. But GenAI has a bigger problem to solve. It is the same one that my management students faced—an inability to differentiate one’s outputs.
The title of a book by Jack Trout and Steve Rivkin, Differentiate or Die: Survival in Our Era of Killer Competition, captures the core belief of every advertiser. All advertisers worth their salt know that the success of a brand depends on its ability to stand apart from its competition.
So, how good is GenAI at creating differentiated output?
Every small step that GenAI takes is determined by the vast data-sets it is trained on. It works by taking its next step based on what past data shows has the highest probability of being next. So, each step GenAI takes is the most predictable, which also means it is the least differentiated.
Much like Ford Motor Company, which once offered cars of “any colour you like as long as it’s black," the modern assembly lines of GenAI are generating undifferentiated output at scale.
There is a bigger hurdle that GenAI will have to cross before it can make its presence felt strongly in the advertising industry. Many think that creating effective advertising is all about generating an image and putting the rational response one wants from the target audience into its headline.
So, if you are creating an ad for a smartphone, in this view, it is about putting out an image of the device with a headline shouting, “Smartphone with XYZ pixel camera," or some other such benefit. If this is advertising, GenAI should be able generate these in no time.
But this is not how effective ads are made. The advertising development process is not as straightforward as it looks. There aren’t many who understand what it takes to make a great advertisement.
Good advertising is all about identifying the right stimulus, ideally an emotional one, that generates an appropriate response in the target’s mind.
Also read: Advertisements crafted by AI needn’t be short of consumer insights
Take for example, the long-running “Shot on iPhone" campaign run by Apple Inc globally. It had a special series for the 2024 Paris Olympics that was all about “2036 Hopefuls," spotlighting aspiring athletes from around the world who are all under the age of 10.
The emotionally appealing photographs of this series were shot by the acclaimed sports photographer Walter Iooss Jr. There were no rational claims made in any of the advertisements. Instead, they aspired to strengthen an emotional bond between the potential user and the brand.
For anyone driving past a billboard with such an evocative image, the non-conscious response generated would be along the lines of ‘Apple smartphones have brilliant cameras.’ This is quite different from the rational responses that traditional ad campaigns for smartphones seek to evoke.
Understanding this stimulus-response relationship is what crafting a great advertisement is all about. Only a few people with high levels of creative intelligence can uncover what works best to strengthen this brand relationship. GenAI still has a long way to go before it generates advertisements that are truly creative and effective.
What is the way forward for GenAI to help generate effective advertisements? Consider the deeper insight of the masala pouch in the noodles packet. The pouch is a reminder that human actions add value. It involves us in the preparation’s vital aspect of taste-checking.
Also read: Creative stunner: GenAI works well where accuracy isn’t everything
Likewise, human creativity must accompany the use of any GenAI tools. GenAI might be able to generate images, headlines and make advertisements in a jiffy, much like a pack of instant noodles. But what will work better are ad campaigns that involve human creativity as an integral part of the creative process. That the output of such an approach will be ‘tastier’ is not hard to guess.
