How The Salvation Army used AI to turn ‘out-of-stock’ into in-store traffic.

During the holiday’s , The Salvation Army logo is everywhere. From toy drives to meal assistance, the organization has become a staple of the season.

Last fall, however, its goals looked more like those of a traditional retailer. The brand set out to drive store visits at scale across key U.S. regions and turned to its agency of record, BarkleyOKRP, to make it happen.

As a participant in the Google Creative AI Lighthouse program, BarkleyOKRP was already exploring ways to use AI to create an “impossible ad.” The Salvation Army’s need for relevance at scale presented the ideal opportunity. Using the latest versions of Gemini and the image generator Nano Banana, the team built a system that transformed a counterintuitive idea into a premium, personalized, nationwide retail campaign—all in just 30 days.

Developing the creative concept.

The campaign centered on a simple objective: turning the emotional appeal of thrifting into real-world foot traffic. “Our insight was simple: FOMO is real, and in thrift, every product in our inventory is one of one,” said Tim McCracken, SVP of creative and AI at BarkleyOKRP.

Rather than promoting items currently in stock, the campaign showcased rare pieces that had already been sold. “It turned absence into attraction,” McCracken said, “and introduced a new kind of retail storytelling—where every ad reminded people to act quickly or risk missing their next great find.”

Because the featured products were no longer available, AI wasn’t just a tool for efficiency—it was essential to bringing the idea to life.

Hyperlocal execution in weeks, powered by AI.

With a bold concept in place, BarkleyOKRP’s team of creatives, developers, and technologists focused on how AI could accelerate execution. Faced with thousands of product photos from The Salvation Army’s nationwide inventory, manual review wasn’t feasible. Instead, they used Gemini to build a system capable of analyzing the entire dataset.

By integrating Nano Banana into that system, the team rapidly curated products and generated creative. Hundreds of stylized, editorial-quality images were produced from basic stock photography, then resized instantly for distribution. To deliver these assets to the right audiences, the team used Demand Gen, a discovery-driven campaign solution designed for visual storytelling.

Gemini processed logistical data such as sale locations and store coordinates, while Demand Gen enabled localized delivery. By targeting ZIP codes near stores, each ad featured imagery tailored to its specific area, allowing the campaign to resonate across diverse markets.

McCracken noted the elevated quality of the final creative. “[Nano Banana] transformed raw inventory photos into beautifully art-directed images that felt like high-end fashion editorials,” he said. “We produced some of the most polished, high-craft work we’ve ever created for the brand.” He also highlighted the tool’s precision in replicating each item, capturing not only color but texture, fabric, and movement.

“It raised the bar for what AI-assisted creative can be,” he added. “This wasn’t just a production tool—it was a true creative enabler.”

Driving ROI, one click at a time.

Within the first 30 days, the campaign demonstrated how a strong emotional insight—combined with generative AI—can drive real results. It delivered an $11 cost per store visit, outperforming industry benchmarks by 138%. Its click-through rate exceeded Google Display Network benchmarks by nearly 2.6X. Perhaps most notably, more than 58% of clicks led to in-store searches, indicating clear intent to visit.

“The campaign took a counterintuitive idea—advertising items that are already gone—and turned it into a performance success story,” McCracken said.

The hyperlocal Google Display Network campaign delivered measurable in-store impact during its initial run, prompting BarkleyOKRP to explore scaling the AI system behind it.

Looking ahead, McCracken sees even greater potential as the technology evolves. If developed today, he believes the campaign could achieve an even higher level of personalization.

His advice to other creatives: treat AI as a multiplier, not a shortcut, and maintain ownership of the work.

“Start with the data and assets you already have, then identify how AI can surface the stories within them,” he said. “The most effective AI-powered work is driven by insight and iteration—not just speed.”

Turn your limitations into your next advantage. Contact Chief Growth Officer Jason Parks, at jparks@barkleyokrp.com to learn more.