The average knowledge worker loses 2.5 hours per day searching for information, reformatting documents, and drafting the same types of text over and over. That's roughly 600 hours per year — gone. Nestlé, the world's largest food and beverage company with brands from KitKat to Nescafé, decided to do something about it. The result was NesGPT: a secure, enterprise-grade generative AI assistant now used by more than 100,000 employees globally. One year after launch, those employees report saving 45 minutes every single week. That's not a rounding error — it's a structural shift in how one of the world's most complex organizations operates.
Most companies are still debating whether to pilot AI tools. Nestlé piloted, scaled, and measured — and the data is in. While competitors were writing AI governance policies, Nestlé was generating 230,000 prompts in a single quarter from its U.S. offices alone. The gap between early movers and late adopters in enterprise AI is no longer a question of "if" — it's already widening.
The story of NesGPT is not just about a chatbot deployed at scale. It's a case study in how to introduce AI into a complex multinational organization without chaos, shadow AI, or a single embarrassing leak. And the numbers back it up.
Why Nestlé Built Its Own AI Tool Instead of Using ChatGPT Directly
By early 2023, Nestlé's IT teams were noticing something uncomfortable: employees were already using consumer-grade AI tools — ChatGPT, Bard, and others — to get their work done faster. This phenomenon, known as Shadow AI, poses serious data security risks: proprietary recipes, unreleased product data, customer contracts, and internal financials could be entered into systems with no enterprise data protections.
Nestlé's response was not to ban AI use — that battle was already lost. Instead, the company partnered with Microsoft to build NesGPT: a private, secure deployment of the same underlying technology that powers ChatGPT, hosted on Microsoft Azure OpenAI Service. Employees get the productivity benefit of generative AI; the company keeps its data inside its own controlled environment. The pilot launched globally in May 2023. By August 2023, NesGPT was live in North American offices. Today, it is available to Nestlé employees worldwide.
"Our goal is to help lighten the weight of work for employees by streamlining processes and empowering them with access to more on-demand information," said Shan Collins, Head of IT North America at Nestlé. "So they can free up time to do their best thinking."
What NesGPT Actually Does Across the Business
NesGPT is not a single-purpose tool. Across Nestlé's sales, marketing, legal, HR, and product innovation teams, it handles a wide range of cognitive tasks that previously required significant manual effort:
| Business Function | NesGPT Use Case | Estimated Time Saved |
|---|---|---|
| Sales | Drafting customer proposals, summarizing call notes, pricing analysis | 30–45 min/week |
| Marketing | Generating copy variants, proofreading, briefing documents | 45–60 min/week |
| Legal | Summarizing contract clauses, draft initial responses | 60+ min/week |
| Product Innovation | Concept generation, market trend analysis, consumer insight synthesis | Weeks per project cycle |
| HR & Operations | Meeting agendas, internal communications, policy explanations | 30–45 min/week |
The variety of use cases is itself a finding. When a tool is genuinely useful, employees find applications their IT teams never anticipated. Nestlé's training program — which focused on how to write effective prompts rather than dictating specific use cases — was designed precisely to encourage this organic adoption.
The Results After One Year: What the Data Shows
In a 2024 internal survey of NesGPT users across U.S. offices, Nestlé found that employees were saving an average of 45 minutes per week. That figure is more significant than it sounds. Across 7,000 active U.S. users alone, 45 minutes per week equals approximately 262,500 hours of reclaimed productive time annually — the equivalent of 131 full-time employees working a full year.
Beyond time savings, employees reported creating better and faster content, and spending less time searching for internal information. In just three months, those 7,000 U.S. users generated nearly 230,000 prompts — an average of roughly 11 prompts per user per week, suggesting the tool had moved from "experimental curiosity" to daily work habit.
For product innovation specifically, the impact was even more dramatic. Nestlé's proprietary Gen AI innovation tool — built on top of NesGPT's infrastructure and connected to consumer insight data from more than 20 Nestlé USA brands — accelerated the product ideation process from six months to six weeks. A process that once required months of manual market research, brainstorming sessions, and concept development can now be seeded with AI-generated concepts in just over a minute.
"Innovation is part of Nestlé's DNA, and our investments in new tools and capabilities like AI are significant and central to fueling our growth," said Veeral Shah, Chief Digital and eCommerce Officer at Nestlé USA.
How Nestlé Drove Adoption Without Mandating It
One of the most instructive aspects of the NesGPT rollout is what Nestlé did not do: they did not issue a company-wide mandate requiring everyone to use the tool by a deadline. That approach typically produces compliance theatre — people logging in to avoid penalties, not using the tool meaningfully.
Instead, Nestlé used a training-first, culture-second strategy. After launch, they ran multiple training sessions across the organization focused on prompt writing — the skill that determines whether an AI tool is useful or frustrating. They communicated ongoing tips and real-world use cases to employees throughout the year. They let early adopters spread enthusiasm organically. And they measured: by tracking prompt volume and conducting user surveys, they identified what was working and refined the program accordingly.
The result is a tool with genuine, measurable traction — not inflated adoption metrics driven by mandates. "Since its initial rollout, we've provided several training sessions across our organization while also continuously working to develop our employees' broader digital capabilities," Collins noted.
The Lesson Other Companies Are Slow to Learn
According to McKinsey's 2024 State of AI Survey, 72% of organizations were using AI in at least one business function — up from around 50% over the previous six years. That sounds like progress. But "using AI in at least one function" often means a handful of experiments in a single department, not the kind of embedded, measured, company-wide deployment that Nestlé has achieved.
Companies that moved early — as Nestlé did in mid-2023 — now have a structural advantage: their employees are more experienced with AI tools, their AI governance frameworks are tested and refined, and their AI-generated productivity gains are compounding. Organizations that are still in "pilot mode" in 2025 are not just behind — they are letting the gap widen with every quarter.
The cost of inaction is not hypothetical. Every week a competitor's marketing team uses AI to generate ten copy variants while yours generates one manually is a week of compounding disadvantage. Every product ideation cycle that takes six months while a competitor's takes six weeks is a gap in speed-to-market that accumulates.
Frequently Asked Questions
Is NesGPT the same as ChatGPT?
NesGPT is powered by the same underlying model technology as ChatGPT (via Microsoft Azure OpenAI Service), but it is a private, secure deployment hosted within Nestlé's own controlled environment. Employee data and company information do not leave Nestlé's infrastructure — which is the fundamental difference from using ChatGPT directly.
How long did it take Nestlé to see results?
The global pilot launched in May 2023 and rolled out to North American offices in August 2023. By July 2024 — roughly one year in — Nestlé published internal survey data showing 45 minutes saved per week per user and 230,000 prompts generated in three months. Meaningful results at scale emerged within about 6–12 months of launch.
Can smaller companies replicate what Nestlé did with NesGPT?
Yes — the underlying technology (Azure OpenAI Service, OpenAI's API) is available to organizations of any size. Microsoft's Copilot for Microsoft 365 provides a similar enterprise-grade deployment for companies already on the Microsoft stack. The key success factors — prompt training, measured adoption, a culture-first rollout strategy — are equally applicable to a 50-person company as to a 270,000-person multinational.
The tools Nestlé used to build NesGPT are already commercially available. What Nestlé added was organizational intent, a training program, and the discipline to measure outcomes. The technology is not the constraint — the decision to start is.