With the advent of generative AI, a lot of my work involves taking consumer insights and using gen AI to help clients design new products and services, or showing them how to use the technology to engage customers in meaningful ways. Traditionally, we’ve had content, marketing, and advertising designed for consumption by the masses, so everyone sees the same thing. But gen AI allows us to move into a world where content and advertising is tailored to the individual, through the insights we’ve collected.
Take car shopping, for example. Let’s say I care more about the seats and the feeling when I'm driving the car, while someone else cares more about the engine and its power. Since gen AI allows automotive companies to tailor its messages to each potential customer, the messages I see focus on luxury, comfort, and seating, while the other person sees content highlighting engine power. With gen AI, you go from a world of everyone seeing the same thing to one where everyone sees very specific messaging tailored to what they care about.
One important risk to bear in mind for companies using gen AI to engage with customers is ensuring your models don’t use consumer data and insights in a way that feels creepy to them. For example, say I'm a car company and have data about the specific attributes of a car that a customer finds attractive, like leather seats. I could very easily use gen AI to create messages focusing on leather seats in a way that alerts the customer to the fact I know this about them.
The really important thing is putting the right guardrails in place. That means developing guardrails to ensure the content you're creating is safe, devoid of toxicity and bad messaging, and considers the customers and how they react to that content. Since you will be sending millions and millions of messages, make sure the models you design and the processes you employ take these into account when using Gen AI for content creation.
