Hershey accelerates confectionery innovation with generative AI

Hershey has integrated generative artificial intelligence into its product development and marketing processes, cutting development times and revealing new flavour ideas, marking a significant shift in the confectionery industry.

The Hershey Company has quietly retooled how it invents and markets confectionery, deploying generative artificial intelligence to compress development schedules, surface novel flavour ideas and tune advertising where it matters. According to a...

case prepared by Greg Kihlström for Sage Business Cases, Hershey has layered external generative platforms such as Ai Palette and ZS onto its product-innovation process, cutting concept-to-market timelines that once stretched for months down to a matter of weeks and rapidly synthesising sprawling consumer datasets to reveal unexpected opportunities , the case cites examples including flavour concepts like a Blueberry Muffin Kit Kat.

The shift is both technical and organisational. Hershey’s digital investments extend into the factory, where the company’s Manufacturing Technology Foundation and a so-called Digital Lean programme aim to create a resilient, data-led backbone for operations. The company said in a blog post that digitising daily management and embedding machine-learning tools has given teams real‑time visibility and ownership, enabling faster iteration and consistent quality as new concepts move toward scale.

AI’s impact at Hershey has not been limited to ideation. According to Adweek, the business used custom algorithms to optimise media placements during a Halloween campaign, reallocating advertising spend in near-real time toward underperforming markets and pulling back where products were already selling. That dynamic approach reportedly drove a 90% sell-through rate in targeted, lower-performing regions and strengthened negotiations with retail partners by demonstrating measurable lift tied to media decisions.

External collaborations have multiplied the effect. IDealogic paired Hershey with an AI analysis partner to mine more than 100,000 online conversations and identify relevance gaps among Gen Z consumers. The work produced product concepts such as a Chocolate Donut Dessert bar and the Mooolicious Bar, designed explicitly to resonate with younger taste profiles and cultural cues. Such examples illustrate how social-listening driven by AI can surface fast-moving preferences that traditional ethnography might miss.

Industry consultancies emphasise the systemic gains companies can expect. ZS, one of the platforms Hershey has used, outlines a generative-AI marketing toolkit that trains agents on wide datasets to accelerate idea generation, simulate consumer or competitor behaviours and run early-stage checks that traditionally required multiple teams. ZS argues this can triple the speed of innovation cycles and lower costs by as much as 30% through reduced rework and retesting, a claim that aligns with Hershey’s reported compression of time-to-market.

Hershey’s experience also sits within a broader confectionery and packaged-goods trend of blending algorithmic scale with human creativity. Confectionery News highlights campaigns such as Starburst’s “Different Every Time,” which combined generative assets, augmented-reality activations and music partnerships to create highly personalised experiences. Mondelez International, meanwhile, has sought to couple generative output with empathetic human oversight, working with agencies to produce personalised images, video and text that seek to preserve emotional resonance while scaling reach, as reported by Forbes.

Those examples point to a recurring tension for consumer brands: the need to capture speed and pattern recognition from AI while preserving the cultural and emotional nuance essential to food and snack brands. Hershey’s approach appears to balance that trade-off by using AI to generate and prioritise concepts, then validating them through testing and through the company’s strengthened digital operations before committing to wide‑scale production.

The changes also map neatly to classic diffusion dynamics. Rapid, observable wins , such as swift concept development and measurable uplifts in retail sell‑through , create demonstrable relative advantage. Integration with existing manufacturing and marketing routines through the Digital Lean initiative increases compatibility, while partnerships and internal communication channels have helped make early successes visible and contagious across the organisation.

Risks and limitations remain. The case and accompanying industry pieces underscore that algorithmic suggestions require careful curation to avoid cultural misfires and regulatory or quality oversights; consultancies caution that generative agents are tools that should augment, not replace, cross-functional judgement. Hershey’s public statements frame the technology as an enabler of human decision-making rather than an autonomous creative director.

As confectionery firms navigate rapid change, Hershey’s story suggests a pragmatic blueprint: invest in data and factory digitisation to make the organisation receptive to AI insights, use external platforms and partners to expand creative reach, and anchor fast experimentation with measurable retail outcomes. Those steps have, by Hershey’s account and by industry reporting, turned generative AI from an exploratory curiosity into a driver of faster product cycles and more targeted marketing , while leaving final editorial control and consumer-facing nuance in human hands.

Source: Noah Wire Services