The opportunity at a glance
As part of its modern workplace and AI strategy, Tower set out to explore how Microsoft 365 Copilot could improve productivity and quality of work, while maintaining strong governance and data security controls. Arinco delivered a security-first engagement combining Data Security Envisioning with a phased Copilot enablement pilot. Within the first month of the pilot, the 100-person cohort achieved full active adoption. Tower users created nearly 300 agents and identified more than 1,000 hours of productivity uplift. Tower gained a clear, data-backed pathway to scale Copilot confidently, securely and responsibly across the organisation.
About Tower
Tower is a New Zealand-based general insurer headquartered in Auckland, serving more than 300,000 customers across New Zealand and the Pacific. Founded in 1869, Tower is New Zealand’s only Kiwi-owned and operated general insurer, offering car, home, contents, business and specialty insurance products through direct and partner channels.
Business challenge
Tower recognised the potential of Microsoft 365 (M365) Copilot to enhance internal productivity, improve quality of work and streamline day-to-day operations. However, leadership was clear that AI adoption needed to be deliberate, governed and secure.
Like many organisations, employees were already experimenting with publicly available AI tools. While this demonstrated curiosity and innovation, it also highlighted the need for a secure, enterprise-grade solution with appropriate oversight, auditability and data controls.
Before enabling Copilot more broadly, leadership needed clear visibility into its data posture to ensure AI could be introduced without unintended exposure. Tower aimed to build internal capability, explore governed agent use cases, and generate measurable evidence of value to inform future investment decisions.
Solution
Arinco partnered with Tower to deliver a structured, governance-led engagement to enable M365 Copilot securely, confidently and with measurable impact.
The program was delivered across two connected workstreams:
- Data Security Envisioning
- Phased Copilot enablement pilot
The program ran over several months, with an intensive four-week pilot phase involving 100 users across 22 teams.
Data security and AI readiness
Arinco began with a Data Security Envisioning assessment to provide Tower with a clear, evidence-based view of its M365 data landscape and AI readiness.
Real-time discovery across Exchange, SharePoint, OneDrive and Teams identified oversharing risks, stale content and permission structures that could create unintended exposure through Copilot. The assessment revealed sensitive and legacy data requiring remediation, along with governance gaps to address before scaling AI. Early generative AI usage patterns also reinforced the need for stronger oversight.
Arinco translated these insights into prioritised recommendations aligned to Microsoft Purview and SharePoint governance controls, enabling Tower to remediate risks and establish a Copilot-ready security framework.
This phase shifted Tower from informal experimentation to a secure, enterprise-grade AI environment with appropriate safeguards and auditability.
Copilot enablement and adoption
With robust security and governance as the foundation, Arinco rolled out a carefully structured four-week Copilot pilot at Tower, engaging around 100 users across diverse business functions. The enablement was phased and tailored to specific audiences, building practical capability aligned to real work.
Tower achieved 100% active adoption within the first month, with Copilot seamlessly integrated into daily workflows. The program combined executive briefings, hands-on workshops, and an adoption hub with open Q&A, accelerating user confidence and practical experimentation. In the spirit of Michelle Hewlett’s analogy, prompting with Copilot was compared to playing the piano: the more you practice, the better you get. As users refined their skills, they found that frequent engagement led to greater fluency and effectiveness, much like mastering an instrument.
Arinco configured analytics and tracked usage, productivity gains, and user sentiment against agreed success measures defined with Tower. Feedback was overwhelmingly positive, with users reporting frequent engagement with Copilot and high scores for quality (8.5/10), speed (8.5/10), and productivity (8.5/10).
User feedback reflected strong enthusiasm for Copilot. Pilot participants shared comments such as, “I think I’d be lost without it now,” “It’s saving me heaps of time,” and “Copilot = velocity + quality.” Real-world use cases ranged from meeting transcription and action extraction to rapid email and document drafting in Tower’s tone, fast retrieval of policies and correspondence, and early agent pilots such as a marketing tone-of-voice agent and an executive board persona agent for meeting prep. The examples were captured and will feed into the next part of Tower’s journey.
By the end of the pilot, Tower was operationally and culturally prepared to scale Copilot with confidence.
Outcomes
Tower completed the Copilot pilot with strong adoption, measurable productivity gains and clear evidence for broader rollout. Key outcomes included:
- 100% pilot adoption: All 100 users embedded Copilot into daily or weekly workflows, with rising prompt volume signalling sustained, habitual use.
- 1,523 hours saved: This productivity uplift was recorded during the four-week pilot, equating to approximately $76.15K NZD in quantified value.
- Nearly 300 agents created: Users developed close to 300 agents supporting practical use cases, generating 478 recorded interactions.
- High user confidence: Feedback scores averaged 4–4.5 out of 5, reflecting growing trust and capability.
- Governance and analytics embedded: Strengthened controls and new adoption reporting frameworks provide leadership with ongoing visibility into usage, value and engagement.
- Customer-aligned scale readiness: The pilot improved internal efficiency and decision-support while establishing the governance and capability foundations to scale Copilot responsibly and enhance customer experience.
What’s next
Following the successful pilot, Tower is progressing formal business case development for Microsoft 365 Copilot, using strong adoption signals, quantified productivity gains and user feedback to guide the next phase of investment. With Copilot now proven in day‑to‑day work, the focus is shifting from pilot enablement to structured scale, deeper scenarios and long‑term value creation.
The next phase will expand Copilot enablement across all business units through Wave 2 rollouts, with a targeted focus on leaders and priority personas. Learning will build on real pilot examples, combining general productivity scenarios with more advanced applications aligned to how teams work. This approach will accelerate value while maintaining confidence and consistency, building on the pilot’s momentum.
Tower is also advancing its agent strategy, moving from early experimentation and personal productivity agents, into a more deliberate, enterprise‑ready model. “Hero” agents will be developed and deployed to all users to demonstrate impact across the organisation, while laying the foundations for a managed agent portfolio over time.
To support this evolution, Tower is strengthening the technical and governance foundations around Copilot and agents. This includes extending data readiness activities, establishing a Power Platform landing zone for agent development and lifecycle management, and beginning to enable Agentic DevOps to bring greater discipline, automation and safe delivery to AI solutions.
With strong governance foundations, an engaged user community and a clear roadmap for Copilot and broader AI innovation, Tower is well positioned to scale responsibly and continue its AI journey with confidence, in partnership with Arinco.
