CPG’s AI moment has arrived

Any big-name CPG firm you can think of is likely to have made at least one AI move already. Autonomous price optimizers, dynamic shelf allocators, and generative brand storytellers are all becoming part of the CPG crowd. They are not all humans – they are AI agents that can contextualize information, plan, and act on it instantaneously.

These tools are helping to create a world where everything – pack design, promotion, shelf layout, even portfolio decisions can be sensed, simulated, and served to every customer moment-by-moment, and at a speed and scale that wasn’t possible before. Here are some examples.

Supply chains don’t just move; they think

Stockouts, pricing errors, promotional errors, and last-mile issues bring costly sales losses. In North America alone in 2024, these issues cost CPG companies $60 billion. That’s why companies are using AI to create more intelligent, dynamic supply chains that can respond quickly to changing trends.

At Mars, the Atlas AI demand-planning suite has unified forecasts across 60 countries, replacing hundreds of spreadsheets with a single AI report. It is wired into the company’s new $450 million smart factory for Royal Canin pet foods, where it flexes production the instant there is a spike in demand.

Content speed beats media spend

Social media and streaming services move at lightning speed. GenAI can help CPG companies to keep pace with consumers’ moods – look at what Unilever is achieving as it partners with thousands of social media influencers.

Unilever’s Beauty AI Studio combines AI with physics, data, and material science to create digital twins of its products. Those twins are then fed into the company’s Gen AI Content Studios and Sketch Pro design unit to generate social-first creative content.

The system can spin hundreds of influencer-ready assets for a brand in under 48 hours – twice as fast as traditional methods, at 50% of the cost. A recent campaign for a Dove limited edition racked up 3.5 billion impressions at a fraction of the cost of TV advertising.

Digital twins erase the cost curve

Generating content for always-on feeds can also prove expensive – until you put GenAI on the job. Nestlé is combining digital twins of its products with an AI-powered content service, to stay ahead of the game. The system can instantly tailor product assets for different platforms or audiences, without the need for manual effort. It speeds up time to market for seasonal or localized campaigns, while enhancing consistent branding across markets and channels.

Nestle’s digital twins also connect directly to planograms, shopper data, and retail testing. This means that the company can rapidly roll out virtual shelf tests and better position its brands in response to consumer behavior.

Some 10,000 product twins will be created by 2027, but the ones in use today have already cut the company’s e-commerce imagery and shelf-test costs by around 70% - making “design once, localize infinitely” a reality.

Human X Machine = Exceptional

We predict that AI agents will outnumber humans in the enterprise by 2027 – so where does that leave humans? The truth is, AI cannot replace human ingenuity. Its true value lies in augmenting it.

The future is not humans or machines. It is humans x machines.

Picture the scene:

• At the breakfast brainstorm, a strategist frames a new purpose for a snack brand – and by 10 AM, an AI model has drafted 100 culturally-tuned storyboards.

• At the mid-morning shelf check, a planner spots an out-of-stock alert in Lagos – so digital twins reroute the inventory before the first afternoon shopper arrives.

• During an afternoon R&D sprint, food scientists sketch a new plant-based meal idea – and GenAI simulations test 1,000 ingredient permutations overnight.

Consumers aren’t the only winners in this scenario. With AI agents on the team, people can give their creativity free reign.

Here’s a real-life example. Procter & Gamble is proving the power of humans x machines with Project Genie, its company-wide GenAI program, which harmonizes point-of-sale, media, and supply data in one place. It has already helped to reduce ad-testing costs by around 90% – but the company wanted to test its value in human, as well as monetary terms.

To do that, it took part in a field study run by Harvard Business School. In a one-day virtual workshop, Procter & Gamble employees were tasked with developing solutions to solve real business challenges facing the company’s brands. The results were impressive. Teams equipped with the data-fed GenAI assistant didn’t just work 12% faster than control groups; they also generated more “exceptional” ideas than either humans or AI alone.

But that wasn’t all. The study found that AI helped employees with different backgrounds to develop more balanced solutions, regardless of their individual expertise. And on a socio-emotional level, AI helped to elevate employee experience and boost morale.