Unearthing AI’s Real Potential to Marketers Requires Human Thinking and Creativity

As we dove into a recent project with the end goal of informing a client’s website optimization, we knew that critical to the task at hand was getting inside the mind of our many complex targets using CX methodology and design thinking. There were personas and journey maps to develop that would solve pain points. Working within Agile project delivery meant critical timelines and estimates. A daunting task to involve AI? If done smartly, actually, no.

As we work through how we—and in this case, ChatGPT-4—can benefit our clients, we reiterate a mantra that we’ve followed in our 40-year history that has never rung truer than with AI: A better brief begets a better result.

Kicking off our research-oriented project meant defining parameters, steering AI in the right direction, and being specific, including giving guidance on tone and providing references. That last point being an important one. Working with AI when your scope of subject knowledge is limited is not unlike crafting a creative brief for a brand through guesswork—hit or miss at best, going down an irrelevant rabbit hole at worst. Fortunately, BNO’s broad experience in the world of healthcare, from pharma to medtech, helped us jumpstart our AI journey, steering it in the right direction including accurate job titles and departments to help feed the research.

Once upon a time there was a global system used to organize libraries called the Dewey Decimal System and it took someone with a master’s degree in library science to help us conduct thorough research. Then came organic search and the URL. Think of the public’s access to AI as a similar leap from the early days of the web. Just like back in the day, you still need to start with smart—or at least informed—questions and a good idea of what you need answered. Bring good foundational knowledge to inform what you’ll ask of AI, and you should be assured of a solid start.

As we worked AI through our project needs, filling in details for complex targets and giving keywords and sources it could tap into, we could eventually begin to ask more and more. We could challenge it to come up with detailed lists for each of the positions for which we were developing personas. Directives like “Give us a list of the most critical administrative challenges facing a (fill in your specialized healthcare role) in 2023” serve up an idea of the strategic breadth and depth that AI can delve if given a good starting point. And lest you think we spent an inordinate amount of time in this upfront AI briefing, the BNO crew working on the project concurred that it would have taken about five times the time investment without AI’s help.

When we’ve tried exercises using AI app Midjourney as a somewhat simplistic replacement for a typical agency assignment—such as coming up with an ad campaign for a product—the results were flat to mixed. No surprise because AI apps are not set-it-and-forget-it tools; the more you ask of them, the more you need to give. But when you do provide more fleshed-out input, more complex search and data-oriented assignments, we were pleasantly surprised by the results. This synthesis of human and AI training and feedback helped Midjourney demonstrate the diversity of our sample concept; actual human diversity as well as different iterations of the concept to show range. And those ranges are something we’re finding AI can be quite good at.

Despite being on the forefront of what will eventually become a more ubiquitous technology, AI currently needs guidance similar to what our examples provided. Constructive, intuitive feedback makes us better learners and thinkers. AI and machine learning have been hardwired in similar ways. Not nearly as complex as the human brain, of course, but ChatGPT-4 acknowledges and synthesizes materials in somewhat similar ways, just very quickly. More feedback will exponentially improve its performance.

The same goes for tools aside from ChatGPT-4 for text that we’ve been trying out, like Adobe Firefly for image generation, Synthesia for talking-head videos, Toonly for animations, and Daydrm for creative briefs, just to name a few. Like any early adoption, a word to the wise: Bring your agency along to bring some heat, heart, soul, in a word—brand—to the party. Otherwise, you might find not-ready-for-prime-time output.

Takeaways and what’s next.

The #1 thing that both AI and humans need to perform at their best: each other.

There’s no doubt in our minds that AI can partner incredibly effectively with its human counterparts. Like every life-changing innovation before it, the right tool in the right human hands can transform the way we live and work.

Let’s talk: human to human.

Contact us and let’s talk about how BNO, AI, and your brand—in collaboration with our range of services—can take insights to inspiration and translate them to memorable, measurable outcomes.