1st BATCH OF FILES AVAILABLE NOW!

Transactional data has been a big unlock for CPG marketers who're able to amass from DTC and other channels a trove of individualized consumer data that can guide their marketing activities. With a big assist from AI, the time now has arrived where non-transactional data like people's social media discussions is underpinning the next wave of targeted marketing. That was the word at Future Smarts from Drew Panayiotou, who took CMO reins at Keurig Dr Pepper a year ago (BBI, Nov 13, 2024). Advertising Age has likened his impact there to a "whirlwind" (BBI, Sep 15). Tho Drew has had runs at consumer brands like Coca-Cola and Chick-fil-A, he described a subsequent move to Google as "my second MBA" for the way it immersed him in the possibilities of AI. At Beverage Digest's Future Smarts conference in NY on Mon, owner/editor Duane Stanford gently coaxed speakers from Big 3 soft drink players toward discussion of how they're weaving AI discipline into their varied activities. But Panayiotou, whose hiring announcement specifically listed his AI prowess, didn't seem to need any coaxing. After spending his first year at KDP laying foundation, that wave is now happening, he promised.What does the harvesting of non-transactional data mean in practical terms? As an example, the Queens-raised Panayiotou noted the winning of Heisman Trophy that very day by Indiana Hoosiers QB Fernando Mendoza. As it happens, Indianapolis happens to be key emerging market for flagship Dr Pepper brand, along with Dayton, Ohio, and Charlotte, where KDP hopes to eventually get brand to the 20 share it enjoys in Texas stronghold. With AI-assisted monitoring of online conversations, KDP marketers can carefully target its messaging in that market behind brand that's heavily tied its identity to college football. "There are people in Indianapolis who care about it and others who don&#39;t care," CMO said of Heisman nod. The not-caring segment might instead receive messaging about brand's tuition program. "There are probably 8 stories we can run in Indianapolis today and we know which of 8 cohorts to direct them to," he informed.Tho we've used word "target," Panayiotou was careful to note that use of tools like Epsilon engine is more about consumer learning than targeting. "No one wants to be targeted," he said. But when you put that tool to use in learning about your consumer, "you jump the track." (He presumably meant that in positive way.) In case of Dr Pepper, with its cult following, the focus is on creating more "raving fans" for brand, much as marketer had done during his run at Chick-fil-A, which has such a large cadre of ravers it doesn't need to use coupons and other come-ons. "These are the fans who recruit other people" far more effectively than paid influencers.Part of the tumult that Panayiotou has created at KDP has stemmed from influx of key execs from outside bevs, or even CPG. "It's been fun getting back into the beverage industry, but we've recruited from outside," he told Future Smarts audience. After all, if you want to be a disruptor, you shouldn't rely on people who've been following established playbook. So new hires have included likes of Oracle and The Trade Desk vet Ben Sylvan as svp for connected media.Tho bevcos' AI-crafted videos - including some highly visible snafus over past year - have made the headlines, "that's not the hard stuff," Drew advised listeners. "AI, simply put, is about building better brands." To do this, his team has been creating Chat GPT-like agents for every brand, powered by all the transactional and non-transactional data KDP is accumulating. The agents are tapped for opinions on new concepts, from packaging to ads. On one brand that's currently nearing launch, agent was shown 5 packaging options and asked which best communicate the concept. "You get very sophisticated answers and it's quick," in real time, CMO said. Accuracy is at 97% now and "by next year it will be 100%."That "precision model" enabled by AI has been put to use behind long-languishing Snapple brand too, starting in its Manhattan birthplace a few months ago. (Tho he didn't discuss details of program, it's included limited-time return of widemouth glass bottle that had long defined brand as well as popup store.) Tho Snapple entered test on a negative trajectory, initiative put product into well over 1 mil households over 6-wk period, generating significant new household penetration for first time in a long time. "That's the power of the precision model," he said. Now that program will be replicated in other markets, with phase 2 due around Mar.That model also is being deployed on another growth-challenged brand, Keurig itself. Currently there are 13K individual pieces of communication on Keurig "and we're seeing progress there," Panayiotou reported.Asked by his host Stanford to address a key issue of the day at KDP - its controversial plan to acquire JDE Peet's, then split merged companies into separate hot-bev and cold-bev companies - Panayiotou's response put an AI lens on that too. "Focus matters a lot . . . incredibly more in the future, when you ask, how do I put AI in the service of this business?" he said. "When you build an AI model for a business, you have to make it very specific to that business." Hot-bev and cold-bev segments are "fundamentally different" and "we're building fit-for-purpose models" in marketing. If you don't build those, he warned, then you're doomed to get "average growth algorithms."

FUTURE SMARTS: How CMO Panayiotou Is Unlocking Power of Non-Transactional Data at KDP; Wielding 'Precision Model' on Dr Pepper, Snapple, Keurig Brands

Publishing Info