Transport for Greater Manchester
Transport for Greater Manchester boosts productivity with Microsoft 365 Copilot
Transport for Greater Manchester (TfGM) is responsible for coordinating public transport and driving mobility strategy across Greater Manchester. As a large public sector organisation supporting millions of journeys each year, TfGM relies heavily on information workers across departments including finance, customer services, highways, and control centre operations.
To support its digital transformation ambitions and modernise workflows, TfGM explored how AI-powered tools, such as Microsoft 365 Copilot, could improve productivity, reduce administrative burden, and enhance service delivery across the organisation. The organisation was also keen to explore the potential accessibility and wellbeing gains that AI could deliver to its staff.
Following a competitive partner selection process led by Microsoft, Phoenix Software was chosen to deliver and support a large-scale Copilot deployment.
Key facts
- Organisation: Transport for Greater Manchester (TfGM)
- Sector: Public sector
- Users: ~1,200 Microsoft 365 Copilot users (office-based staff)
- Active pilot users: 1,041
- IT/delivery team: 50–100 (IS function managing service desk and security)
- Solutions: Microsoft 365 Copilot, Copilot Studio agents, Microsoft Purview (security and compliance)
- Phoenix services: Deployment, adoption and training, governance, discovery workshops, agent development
- Vendor funded engagement: Initial Copilot deployments supported by vendor investment
TfGM faced increasing pressure to do more with existing budgets while maintaining high levels of public service delivery. Key challenges included:
- High volumes of manual, repetitive administrative work across departments
- Demand for efficiency gains without increasing headcount
- Legacy processes (e.g. manual reporting, document creation, email management) slowing teams down
- A need to modernise ways of working using AI and automation
- Inconsistent output quality and time spent on low-value tasks
- A desire to improve employee wellbeing, reducing overtime caused by admin-heavy workloads
Georgia Platt, Senior Delivery Manager at TfGM summarised the organisation’s approach to the Copilot trial,
“Ultimately, we wanted to understand how AI could make people’s day-to-day lives easier at work; whether that meant streamlining repetitive tasks, improving efficiency, or simply helping people get more from their day. We also wanted to explore its potential impact on staff wellbeing. We approached the project with an open mind, recognising that with such a diverse range of roles and responsibilities across the organisation, it would be difficult to define specific success metrics from the outset. They would come later once we had a partner engaged and developed our use cases.”
The Copilot trial was deliberately timed to align with the upcoming renewal of TfGM’s Microsoft Enterprise Agreement, which was due within six months of deployment. The results of the trial would directly inform the organisation’s level of investment in Copilot as part of that renewal. Ultimately, TfGM wanted to determine whether AI could deliver meaningful reductions in effort, improve productivity, and generate measurable business value ahead of making a long-term commitment.
Phoenix designed and delivered a structured, organisation-wide Microsoft 365 Copilot programme focused on adoption, measurable outcomes, and long-term scalability.
The initiative kicked off with a large-scale deployment of approximately 1,200 Copilot licences across all desk-based staff This was executed initially as a strategic four-month pilot, carefully aligned with TfGM’s upcoming Enterprise Agreement (EA) renewal decision.
- Adoption and Change Management: To develop a “Copilot-first” organisational culture, a comprehensive change management strategy was deployed. Users were upskilled via targeted workshops, webinars, and e-learning. The curriculum went beyond technical training to address AI concerns directly through responsible use and governance education.
- Departmental Discovery and Role-Based Approach: Phoenix collaborated closely with individual departments to identify high-value, role-specific use cases rather than generic training. The rollout was highly targeted, as Georgia explains: “Originally, we focused our efforts on the people and finance teams as these were the departments we felt could gain the most efficiencies at the start. Phoenix organised workshops to analyse current processes and propose solutions using Copilot and Copilot agents. The initiative quickly expanded into other departments from there as colleagues began inquiring about the technology.”
- Governance, Security, and Compliance: Data protection was prioritised from the outset by implementing Microsoft Purview and sensitivity labelling. The programme established robust AI governance and responsible use policies, ensuring full compliance with stringent public sector data standards.
- Advanced Agent Development and Automation: Moving beyond standard out-of-the-box features, the programme integrated custom Copilot Studio agents and Power Apps solutions. These targeted high-impact, labour-intensive workflows, specifically optimising reporting, complaints handling, and financial analysis.
Benefits and outcomes
The TfGM Copilot pilot delivered significant, measurable value within the four-month period:
- 93% initial adoption rate across users
- 1,041 active users
- 63,668 Copilot actions completed (measured over a four-week period, and increasing each reporting period)
- 80% weekly usage; 53% daily users
- 6,017 hours of AI-assisted productivity generated in 4 weeks
- Equivalent to 40 full-time employees (FTEs) added capacity
- Users saving 5–6 hours per month, far exceeding the 7-hour annual break-even point
- Top benefits cited among Copilot users:
- Time savings (26%)
- Increased productivity (20%)
- Improved quality of work (17%)
- £2.34 million annual value generated
- £2.07 million net benefit (pilot)
- 9:1 return on investment
- 2,859 meetings automatically summarised: In four weeks, Copilot processed 2,626 hours of meeting content and generated intelligent summaries across 2,859 meetings. Manual minute-taking has been eliminated in departments such as Facilities Management – freeing up hours per week per manager.
- 868 documents drafted using Copilot: Campaign reports that previously took a full working day are now completed 30–50% faster, saving 2.5–4 hours per report.
- 39,194 AI prompts submitted: Drafting documents, summarising emails, and triaging inboxes. Leaders who spend 40–45% of their time on email projected 3–6 hours saved per person per week.
- Control Centre: Reporting reduced from 6 hours to under 1 hour, saving 20–25 hours per week
- Facilities Management: Manual notetaking eliminated, saving 3–6 hours per manager per week
- Customer Services: Complaints handled 30–45% faster
- Finance: 2–3 hours saved per cycle on reporting tasks
- Business Development: Reports produced 30–50% faster
TfGM is now entering the next phase of its AI journey by developing targeted Copilot agents to automate high-impact workflows. These include solutions for network status reporting, complaints resolution, financial bid analysis, and contract knowledge management.
“We aim to unlock even greater efficiencies with Copilot agents, by embedding AI-driven processes into everyday operations and reducing manual effort even further,” concluded Georgia.
Looking to unlock similar productivity gains with Microsoft 365 Copilot?
Contact Phoenix Software to help you design, deploy and scale AI solutions that deliver measurable outcomes in your organisation.