There’s something exhilarating about watching a persona come to life—not on a screen, but from a handful of words, ideas, and visuals woven together by AI.

When we start crafting AI-generated personas for a campaign, the process feels like creating a character for a film. My design approach is rooted in storytelling, and the first step in shaping these personas always begins with the moodboard. I look for textures, colors, fashion, lighting. The character isn’t just about how they look; it’s about the world they inhabit, the emotions they evoke, and the narrative that unfolds through their presence.

Initial prompt: young woman holding cellphone showing big eyes, smiling and laughing, in the style of yellow and lapis blue, pop-culture-infused, kombuchapunk, commercial imagery, interactive experiences, sterling silver highlights, contest winner

Crafting a prompt is a nuanced process, more like directing a photoshoot through words than simply giving instructions. I’m not just asking the AI to create an image—I’m specifying the exact type of imagery, the mood, and even the technical details, like the camera that should be "used" in the AI's "photographic" process. For example, I might want a shallow depth of field to give the image that cinematic quality, or request soft lighting with golden undertones that evoke warmth and nostalgia. Once I hit the right combination of words, the AI generates its first pass—a sketch, a glimpse of the persona’s potential.

Initial prompt: man wearing headphones swings in front of a living room, in the style of fashion photography, dark green and amber, groovy, folk-inspired, sleepycore, brown and bronze, catcore

Rarely does the first output hit the mark. The first render is like a rough draft—there’s potential there, but it needs refining. So, I iterate. I’ll adjust the prompt, maybe add references to specific art styles that inspires the look I’m going for. I might tweak the lighting, rework the angle, or adjust the persona’s expression. Each iteration brings me closer to the direction that feels just right.

And while it isn’t perfect yet, character reference allows us to create the persona, and adjust things like clothes, expressions, and setting.

When I get a version I’m confident in, we move into the next phase: initial retouching. This stage isn’t about perfecting the image yet; it's about making the persona look polished enough to present for internal reviews and client discussions. We refine the image, smooth the edges, and get it ready for the concepting phase, where we show proof of concept with visuals placed in mock ads. This is where we begin to see how this persona interacts with the world we’re building for them—on billboards, in digital banners, across social media, and beyond.

Once the concept is approved, I work with our freelance retouchers and the production team to take the image and version it out into thousands of deliverables. We create high-funnel assets to grab attention, mid-funnel ads to educate and engage, and low-funnel pieces that close the loop and drive conversions. From one AI-generated persona, we build a full ecosystem of creative that adapts to every touchpoint in the customer journey, all while maintaining the integrity and storytelling of the original concept.

It’s this blend of art and technology that excites me most—using AI as a tool to push the boundaries of what’s possible while staying grounded in timeless design principles. Each project is a new opportunity to bring an idea to life in ways that feel fresh, unexpected, and deeply connected to the story we’re telling.