I’m the creative and user experience director for a marketing agency that works with hundreds of CPG and eCommerce brands. I love using AI in my day-to-day job.
Before you start wondering about the likelihood of a creative guy embracing AI, know this: it’s not for producing the actual work. It’s about leveraging AI as a research tool to help me and my team spend more time on original copy and visuals that elicit engagement.
In this article (which, by the way, was not a product of ChatGPT), I’ll spell out my favorite use cases for AI, add a bit about my favorite AI tools (and the ones I won’t trust yet), and throw in a few best practices for helping refine the output.
Let’s get into it.
Use cases for AI in brand and user experience research
I use AI for two main types of research: user research and competitive research.
Let’s start with user research. AI can absolutely help you learn more about your audience. It helps to start with what you know (e.g. if you’re selling jogging strollers, you know your audience is active parents of young kids), but even if you’re starting from scratch, you can ask the tool “What’s the best audience for {product}?” to get a starting point.
Once you get a demo, you can pepper the tool with questions as though it’s a focus group, and the tool will give you generalities on a consensus from the group. In the jogging stroller example, you might ask the tool what the biggest challenge is for a new parent’s exercise routine, and the tool might tell you “finding more time in the day.” You could use that knowledge to craft messaging around how easy the stroller is to use. You can also use AI to ask about adjacent or even very different audiences than your core audience to gauge the prospects of expanding your market.
AI is particularly good for helping with competitive analysis. Once you’ve really locked down details on the business you’re working for (business type, vertical, audience/market), you can ask the tool who the business’s top competitors are. From there, use the tool to drill down into each competitor by asking about their annual revenue, most popular products or services, and differentiators.
My favorite (and least favorite) AI tools
To tackle the above initiatives, the first AI tool I started using seriously (I’m not alone here) was ChatGPT, which is still my first choice for prompt-based research. Recently, I’ve been dabbling with Perplexity for prompt use as well, and I’ve been impressed by the results.
On the other hand, Gemini’s early results were patchy at best and downright offensive at worst. Google’s rush to get it to market didn’t serve anyone (certainly not users, and certainly not Google) – for me, it only locked in ChatGPT as the best option. I’m open to testing anything that comes along, and I make a practice of doing so, but I’ll need to see a ton of improvement in Gemini before I start paying any attention to it going forward.
Best practices for refining AI-produced research
This article won’t go into depth about how to create prompts; that’s worthy of its own topic. In general, I’ll say the more detailed and specific you can make the prompt, the better output you’ll get.
Instead of: “Can you give me information about Brand X’s competitors?”
Try: “Can you provide me with the top 5 competitors in Brand X’s space, their annual revenue and top offering, and a unique quality that differentiates them from their competition?”
You’ll still (and probably always) need to cast a critical eye with plenty of QA on the answers no matter the prompt, but you’ll get richer information from smarter queries. Essentially, I carry the same expectations for AI research as a I would for an entry-level research assistant: they’re not experts, so they lack context; they do what you ask and no more; and ultimately, if you do a little work upfront to give them good direction, you’ll save yourself a boatload of time you can put to good use elsewhere.
Wrapping up
Most creative people I know and respect understand that great creative is built on empathy and understanding of the audience meant to see it. That’s still as true as ever. Using AI to help build a base of that understanding isn’t creepy; it’s a good use of available tools that lets you shift your resources from what machines can help do to what only humans can do – come up with original ideas and continue to break new creative ground.
For more expert articles and industry updates, follow Martech News
ABOUT THE AUTHOR
David Lachowicz, Senior Director of Creative & Development at DMi Partners
David Lachowicz is currently the Sr. Director of Creative and UX at DMi Partners. Since joining the full-service digital agency in 2013, he has worked to develop the company’s creative department to raise the industry standard. His inventive concepts have driven success on a breadth of visionary campaigns designed to deliver engaging experiences, elevate brand awareness, and improve user journeys. David and his team have earned multiple awards for their work from associations including the Addys and the Daveys.