Advertisements crafted by AI needn’t be short of consumer insights

There are a few who think that creating an effective advertisement is as simple as putting the response you want from the consumer into the headline of your ad. So if you want someone to buy a product, advertise “Buy XYZ brand" as the main message. To provide further impetus to buy that product, advertise information on its functional benefits. As the ultimate incentive, mention any price discount on offer too. But the truth is that this is not how effective advertisements are created.

A few decades ago, Jack Trout and Steve Riven, in their book Differentiate or Die: Survival in Our Era of Killer Competition, reminded businesses the importance of differentiating one’s product from others in the market. But this differentiation strategy often has little to do with rational product benefits. The technology and ingredients used to make products in most categories have stayed the same for decades. For example, the ingredients of the toothpaste we use have not changed in the past few decades. Even if a product is technologically superior, very few people can actually discern such technological differences.

Brand differentiation is about generating the right emotional stimulus on behalf of a product. The question ‘Which college do you study?’ in a soap advertisement, for example, could tug the heartstrings of many a young mother in India. To appreciate the persuasion impact of this simple question on the sales of a soap brand, one need to understand the importance of human behaviour insights in crafting effective advertising.

Insights are not peripheral truisms about human behaviour. That demand for a product will go down as its price goes up is not an insight. It is just a truism. Such truisms are often on top of the pile in reports on samples of consumer conversation. But to make differentiated advertisements, such elementary truisms of human behaviour do not help. Knowledge of many young mothers being ill at ease deep within as a result of insecurity brought on by post-pregnancy bodily changes, however, is an example of a behavioural insight. How much ever we churn the past sales data of Wipro’s Santoor soap, this insight will not be discovered. How much ever we churn consumer-conversation data, it would not emerge. An insight is a deep human need that a consumer might not even articulate in a normal conversation. For that matter, on the face of it, one might even deny such thoughts. So such insights may require deep qualitative consumer research to reveal.

Powerful insights come from a deep and creative understanding of consumer behaviour. As part of creating an advertisement for a deodorant in India, my team once had to understand whether the masturbation fantasies of Indian males differed from those in Western countries. What is the data-set that one could analyse to arrive at such an understanding?

Consumers will rarely tell the truth about such deeply personal aspects of their lives. That is when I realized that comparing the editorial content of Indian porn literature and Western porn material may provide an answer to that question. Will GenAI tools ever be this creative in their information search?

If GenAI has to take over any responsibility for brand communication from humans, it should be able to unearth unique human behaviour insights. GenAI platforms are built on Large Language Models with access to far more knowledge than a human being can ever remember. But hoping to get great human behaviour insights from an LLM would be like looking for a needle in a haystack. This is where the role of prompts in a brand’s GenAI strategy takes on huge importance.

In the traditional advertising industry, the creative brief was a well-thought-through document that was prepared after much research and discussions among the strategic minds both at the client’s end and within its advertising agency. A creative brief contained all the important elements needed to make a good advertisement: unique human behaviour insights, brand guidelines, the key response required from the particular campaign or ad, and so on. A close understanding of this brief is what guided human creative teams in developing various creative stimuli to generate the appropriate response from the target consumer.

If GenAI is to take charge of advertising, the crucial role played by the creative brief in the traditional ad industry will have to be taken over by an effective prompt strategy. It will quickly be found that prompting cannot be a rational process managed by tech engineers. The prompt strategy would have to be developed by human behaviour and design experts tasked with getting GenAI to explore the depths of behaviour that qualitative research has not yet reached. The prompt strategy would then be the big differentiator in the brand’s Gen AI game-plan.

Only with a truly creative prompt strategy would GenAI have been able to come up with a line like ‘Which college do you study?’ to sell more soap.