How should leadership be organized to make AI work across the whole company?
To make AI work for everyone, treat it as a company-wide transformation with clear roles across leadership and functions:
- **CIOs**: Define the overall AI strategy, ensure it aligns with business goals, and set the prioritization framework for AI projects.
- **IT leaders**: Own the AI infrastructure, platform choices, and secure deployment. They make sure solutions are scalable, reliable, and compliant.
- **Functional leaders (e.g., sales, finance, HR, operations)**: Organize user adoption teams within their functions and act as sponsors and champions. They identify high-impact workflows and ensure AI is embedded into day-to-day processes.
- **User adoption teams**: Lead AI training, change management, and engagement programs. They help employees understand where AI fits into their work and how to use tools like Microsoft 365 Copilot effectively.
- **Security and compliance teams**: Oversee AI governance, security, and risk management. They evaluate data quality, privacy, regulatory requirements, and acceptable use.
This shared ownership model helps AI move beyond experimentation and into sustained, measurable impact across the organization.
How do we decide which AI projects to prioritize?
A practical way to prioritize AI projects is to use a business-first framework that balances impact, feasibility, and measurable outcomes.
1. **Start with business impact and value**
Ask:
- Does this AI solution address a clear business challenge?
- Can it improve efficiency, revenue, or customer experience?
- Is it scalable across multiple departments or use cases?
Microsoft recommends linking AI investments to at least one of these outcomes:
- **Revenue growth**: e.g., AI-enabled personalization, intelligent sales automation.
- **Cost efficiency**: e.g., process automation, AI-supported forecasting.
- **Risk mitigation**: e.g., AI-powered cybersecurity, fraud detection.
2. **Assess feasibility and readiness**
Evaluate whether:
- You have sufficient data and infrastructure to support the AI model.
- Security and compliance risks are manageable.
- Your organization has (or can build) the skills to implement and maintain the solution.
3. **Define quantifiable financial projections**
Before greenlighting a project, clarify:
- How soon it is expected to deliver value.
- The projected cost savings or revenue gains.
- The KPIs that will be used to track success.
4. **Use a cross-functional AI council**
Microsoft suggests forming a cross-functional AI council, led by a central team, to approve AI projects. Each approved project should have:
- A clear business owner.
- A defined success metric (or set of metrics).
- An agreed scope and timeline.
5. **Focus on high-impact workflows first**
Examples of workflows where generative AI can create tangible value include:
- **Customer service**: support intake, triage, diagnosis, and resolution; customer self-service.
- **Sales**: lead generation, sales engagement, proposals, quote-to-cash, post-sale follow-up and upsell.
- **Finance**: planning and analysis, sales analysis and forecasting, tax and treasury, record-to-report.
- **Marketing**: demand creation, campaign execution, content creation, customer insights and strategy.
- **HR**: recruiting, HR admin and payroll, learning and development, talent management, employee engagement.
- **Legal and compliance**: contracts and agreements, compliance management, risk management.
A data point to keep in mind: through 2025, an estimated **30% of GenAI projects will be abandoned after proof of concept** due to poor data quality, inadequate risk controls, escalating costs, or unclear business value. A disciplined, business-first prioritization approach helps avoid this outcome.
How can we measure AI readiness, adoption, and business impact?
You can structure AI measurement around three dimensions: **readiness, adoption, and impact**, and use tools like Copilot Analytics and the AI Adoption Score to track progress.
1. **Readiness metrics (before rollout)**
These help you understand whether your organization is prepared to adopt AI:
- **Employee sentiment**: Use surveys and feedback tools (e.g., Viva Pulse) to gauge motivation, confidence, and perceived value of AI.
- **Technical eligibility**: Assess whether employees are using the relevant applications today and whether your environment is technically ready for AI deployment.
2. **Adoption metrics (during and after rollout)**
Once AI is deployed, track how it is being used:
- **Usage patterns**: Frequency and intensity of AI-assisted work across teams, apps, and functions.
- **Employee feedback**: Barriers, challenges, and perceived usefulness of AI tools.
- **Collaboration patterns**: Changes in how people communicate and collaborate, which can indicate shifts in work habits.
Microsoft provides an **AI Adoption Score** for Microsoft 365 Copilot:
- A score of **100** means all licensed Copilot users used Copilot features on at least **12 of the past 28 working days**.
- This threshold correlates with sustained adoption and helps you see where additional enablement is needed.
3. **Impact metrics (business outcomes)**
To demonstrate value, connect AI usage to business KPIs:
- **Productivity gains**: Time saved, AI-assisted hours, and efficiency improvements.
- **Business outcomes**: Revenue growth, cost savings, and customer satisfaction.
- **Employee experience**: Changes in satisfaction, engagement, and overall work experience.
Tools like **Copilot Analytics** and the **Copilot Business Impact Report** help you:
- Identify relevant business areas, outcome metrics, and data sources.
- Upload business metrics (e.g., via Viva Insights) and correlate them with Copilot usage.
- Customize analyses by team, metric, and data source to see where AI is driving measurable value.
For example, one organization, Access Holdings Plc., used Microsoft 365 Copilot to:
- Reduce report preparation time from **6 hours to about 45 minutes**.
- Increase staff engagement during meetings by **about 25%**.
- Cut chatbot development timelines from **2–3 months to roughly 10 days**.
A holistic measurement strategy combines behavioral data (usage, adoption) with sentiment data (employee willingness and experience). This gives leaders a clearer view of where AI is working, where it is not, and how to adjust investments and enablement to stay aligned with business goals.
_thumb.jpg)
Sign in with LinkedIn