Vodafone: Copilot for 68,000 employees — ~3 hours saved weekly, while TOBi resolves 70% of ~45M monthly inquiries

• Vodafone to roll out Microsoft 365 Copilot to 68,000 employees to boost productivity, innovation, and quality — Microsoft Source EMEA, 2024-09-16
• Time savings and an enhanced employee experience at Vodafone through its use of Microsoft 365 Copilot — Microsoft Customer Stories, 2024-12-03
• Vodafone amplifies call center innovation, customer service, and employee inclusion with Azure AI — Microsoft Customer Stories, 2025-02-05
• Vodafone and Microsoft sign 10-year strategic partnership to bring generative AI, digital services and the cloud to more than 300 million businesses and consumers — Vodafone (press release), 2024-01-16

Background

Vodafone is one of the world's largest telecom operators: more than 330 million customers across 15 countries in Europe and Africa, and around 100,000 employees. On January 16, 2024, the company signed a 10-year strategic partnership with Microsoft, with $1.5 billion in planned investment in cloud and customer-focused AI services developed jointly with Microsoft. The agreement is broader than "bought Copilot": it includes transforming customer experience with generative AI, hyperscaling Vodafone's IoT platform (spun out as a standalone business by April 2024), new digital and financial services for SMEs across Europe and Africa, and an overhaul of the global data center strategy — all within Vodafone's own responsible AI framework.

Vodafone CTO Scott Petty frames the company's AI strategy in three directions: "Our AI journey is focusing on three areas: operational efficiency inside the organisation; rewiring the business to provide an enhanced customer experience; and unlocking growth opportunities through new products and services."

In practice, the rollout runs on two fronts. Internal — Microsoft 365 Copilot for employees: first a pilot, then a rollout to 68,000 people across all geographies. External — the TOBi customer chatbot, operating in 13 countries and 15 languages, and the SuperAgent assistant for call center agents, both on Azure OpenAI Service. This two-front construction makes the Vodafone case one of the most complete examples of enterprise generative AI in telecom: it covers the back office and customer service alike, and — unusually — comes with an independent impact assessment.

Important context: TOBi existed at Vodafone long before generative AI as a classic virtual assistant. The case describes not building a bot from scratch but reinforcing a mature system with a new generation of models — which distinguishes it from stories that measure impact from a zero base.

Problem

Vodafone's scale makes even small inefficiencies expensive. About 45 million customer inquiries per month flow through contact centers and digital channels. Agents had to manually search 20-page PDF manuals mid-call — every such search stretches the conversation, irritates the customer, and drains the employee. Routine requests — billing, network issues, phone orders — consumed live agents' time even though most don't require a human.

Internally, there are 100,000 employees whose time goes into the same routine: email and agenda drafts, meeting minutes, document summarization, searching for information across internal systems. In the legal department the bottleneck was contracts: drafting, review, renegotiation — document-heavy work where every hour saved is multiplied by the flow of agreements.

The third problem is managerial, and it explains the structure of the case: how do you prove impact measurably before scaling the investment to tens of thousands of licenses? "It feels faster" is not an argument for the board under a 10-year, $1.5B contract. Vodafone needed a methodology: a limited pilot, an external assessor, before/after metrics — and only then a rollout. It is this approach, rather than the specific numbers, that makes the case a model: the company built a process where the scaling decision is made on data, not on early adopters' enthusiasm.

Solution

The internal front began with a disciplined pilot: Microsoft 365 Copilot went to 300 users in an early access programme, with results assessed together with KPMG — an external evaluator with no stake in selling licenses. The pilot scenarios were the most grounded ones: drafts of emails, agendas, and documents; meeting summarization; information search. Based on the results, Vodafone decided to roll out to 68,000 employees across all geographies — within a couple of months of the pilot.

The evolution of usage recorded in the case is telling. Federico Frumento, senior data governance manager: "When it began, it was mostly trying to use it to write a document; like 'write an email for us.' Over time, that evolved into actually using it as a Copilot; using it as an expert sitting by your side." The case also notes the benefit for neurodiverse employees: for people with dyslexia, Copilot's drafts reduce the stress of writing — a rare inclusion angle in enterprise case studies.

The 68,000-person rollout comes with AI integration across a wide range of functions — from customer service and product development to network management, sales, and marketing. Robert Leeson, Lead Digital Workplace, frames the task pragmatically: Copilot should help colleagues "get maximum value from the Microsoft 365 tools they use every day" — that is, embed into the existing Word, Outlook, and Teams rather than become yet another separate app. Petty adds: "Generative AI really does make us more productive and will enable us to have a much greater experience at work."

The legal department became the internal champion: lawyers use Copilot for contract review and summarization. "Microsoft 365 Copilot has helped us to review those documents more quickly," says Hazel Butler of the Legal & Business Integrity team. Measurements showed contract drafting time cut by about an hour per document, and subsequent analysis found average savings of 4 hours per week per person in the legal function.

The customer front is built on Azure OpenAI Service. The TOBi chatbot, operating in 13 countries and 15 languages, was reinforced with generative models: "TOBi can now use Azure OpenAI for context-aware conversations, adapting responses naturally," describes Beverley Bartlett, Head of Digital Care. The bot covers routine requests — billing, network, device orders — and hands complex ones to humans.

For live agents, Vodafone launched SuperAgent — a conversational search interface that surfaces the right information from documentation in seconds, replacing mid-call scrolling through 20-page PDFs. It was piloted at the Corso call center in Italy. The system also includes a quality assurance element evaluating compliance of responses against policies and regulations. The stack: Azure OpenAI Service, Azure AI Foundry, Azure AI Search, Microsoft Copilot. Ahmed Elsayed, CIO UK and Europe Digital Engineering Director: "With Azure AI Foundry, you can interact seamlessly to explore, build, test, and deploy AI tools."

The strategic frame for the two-front program comes from Ignacio Garcia, CIO Italy and Global Director of Data Analytics and AI: "Our mission has always been to connect people and data for a better future. We're still on our journey with AI, but the roadmap is clear."

Result

The Copilot pilot, assessed with KPMG, produced the case's headline number: savings of around three hours per week per person on average — on emails, minutes, and information search. 90% of participants said they benefited and wanted to keep using the tool; 60% said it improved the quality of their work. The legal department later measured an average of 4 hours saved per week per person, with contract drafting time down by about an hour per document. On this data Vodafone decided to roll Copilot out to 68,000 employees — while Petty stresses: "It's not about doing more work, it's about doing better quality work and being more customer-focused." Microsoft's story also records the expectation that savings will grow as users get comfortable with the tool: the pilot's three hours reflect users who had not yet reached "cruising speed" with Copilot.

On the customer front: TOBi handles nearly 45 million inquiries monthly and fully resolves 70% of them through digital channels; the remaining ~30% go to live agents supported by SuperAgent. With SuperAgent, average call time dropped by at least one minute; Microsoft also reports improved customer satisfaction post-implementation — without disclosing specific NPS/CSAT values.

Framing. The three hours is self-assessed by pilot participants in a 300-user trial, albeit collected with KPMG's involvement: it is a survey, not a time study. TOBi's 70% is the metric of a mature system that generative AI enhanced rather than created from scratch: it cannot be attributed wholly to Azure OpenAI. No currency-denominated financial impact has been disclosed on either front; all of the case's sources are Microsoft materials — an interested vendor.

Editorial analysis. The most valuable part of the case is not the numbers but the proof construction: limited pilot → external assessor (KPMG) → scaling decision made on data. That moves the AI-productivity conversation from the genre of "feelings" to auditable metrics — the layer most enterprise Copilot deployments lack. The second value is the SuperAgent pattern: AI for the agent, not instead of the agent. A minute off each call across ~45M monthly inquiries is an enormous operational lever without the service-quality risks of full automation; notably, the sources report no staff reductions tied to these tools. Third: the variance of impact across functions (3 hours on average vs 4 for legal) is a practical argument to prioritize rollout by document-heavy departments rather than spreading licenses evenly. And fourth, a detail easy to miss: the feedback from neurodiverse employees about reduced writing stress shows enterprise AI has an inclusion dimension that never shows up in standard productivity metrics yet directly affects retention and engagement — not for nothing did Microsoft put "employee inclusion" in the title of its Azure AI story.

Lessons learned

1. Independent pilot assessment (KPMG) turns 'it feels faster' into a number you can take to the board — a rare practice that sets this case apart from pure self-reporting.
2. The 300 → 68,000 sequence is exemplary: a measured pilot with control metrics first, scale second — not the other way around.
3. Impact varies by function: legal saves 4 hours vs ~3 on average — prioritize rollout by document-heavy departments.
4. Usage matures over time: from 'write an email' to 'an expert sitting by your side' — evaluate pilots beyond the first weeks or you will underestimate the tool's ceiling.
5. The two fronts reinforce each other: internal productivity (Copilot) and customer service (TOBi/SuperAgent) — different budgets and metrics, one Azure platform.
6. SuperAgent shows an undervalued pattern: AI for the agent, not instead of the agent — a minute off each call across ~45M monthly inquiries scales better than full automation that risks NPS.
7. Self-assessment metrics (90% 'found it useful') and time measurements are different evidence tiers; separate them explicitly in your own business case.

Frequently asked questions