Measuring ROI in AI

The Challenge of Measuring AI ROI

I hear from CEOs frequently about their difficulty in measuring the Return on Investment (ROI) for their AI initiatives. This challenge isn't unique to AI; measuring ROI across all business activities is inherently complex. That's why businesses default to easily quantifiable metrics like Annual Recurring Revenue (ARR) and earnings per share. Ultimately, the core objective is to deliver value to shareholders, but accurately gauging the success of all the contributing elements remains a significant hurdle.

This topic always brings me back to a presentation given by Rudy Cifolelli concerning the alignment of an entire organization when setting and achieving its objectives.

Rudy's presentation is the reason I read the book Measure What Matters by John Doerr. This book introduces the Objectives and Key Results (OKRs) framework, a goal-setting methodology designed to help organizations synchronize their efforts and monitor progress. Doerr uses case studies from organizations like Google and Bono's ONE campaign to illustrate the efficacy of OKRs in driving success.

Why AI Amplifies Existing Problems

Here's the critical insight: AI adoption rapidly highlights weaknesses within existing business processes. Poor data quality, for instance, directly diminishes AI results. Likewise, inadequate security practices can lead to AI accelerating negative consequences. If a process is flawed, AI will amplify those flaws. Furthermore, if your team lacks commitment to their work, AI may become a distraction—like social media—rather than a powerful tool for increased contribution.

This is why measurement matters before implementation. You need to know what you're measuring before you deploy AI at scale.

AI as a Measurement Tool Itself

Here's where it gets interesting: AI, despite its seemingly magical capabilities, is fundamentally just another new technology. Therefore, to accurately measure the Return on Investment (ROI) of a new AI solution, you must first have a system in place for measuring the performance of your teams and the tools they use.

And here's the fascinating paradox: AI can actually be used to develop the very strategy for measuring success within the organization.

Building Your AI Strategy: An Action Plan

A successful AI strategy should begin with complete alignment across your organization on your foundational elements: your core mission, immediate goals, and guiding principles.

Your Strategic Roadmap

• Define and Share the Mission: Clearly communicate your company's mission statement to all employees.
• Foster Alignment: Inspire your team to embrace and align their work with the stated mission.
• Establish Core Principles: Publicly declare your core principles, with a specific and strong emphasis on modern AI essentials like security and data protection.
• Appoint an AI Champion: Designate one individual to serve as the organization's ambassador and driver for all AI initiatives.
• Form an AI Committee: Create a committee whose main purpose is to support and mentor the AI Champion, to whom the Champion will report.
• Crowdsource AI Tool Requests: Encourage your people to submit requests for specific AI tools. Each request must clearly articulate how the tool supports key performance indicators (KPIs) and include a high-level assurance of alignment with the company's compliance and data protection policies.
• Ensure Rigorous Diligence: Empower your cybersecurity and compliance teams to perform their due diligence—this step is critical.
• Implement and Test AI Tools: Roll out AI tools through flexible pilot programs or month-to-month engagements, utilizing multiple providers for testing and comparison.

A Framework for Measuring AI Pilot ROI

To accurately assess the Return on Investment (ROI) for an Artificial Intelligence (AI) pilot, a structured approach is essential:

1. Assign a Steward: A dedicated individual must be appointed to oversee the initiative from the outset.
2. Establish Clear, Achievable Goals: The pilot's objective should be specific and realistic, avoiding overly ambitious targets.

Three Areas for AI Investment and ROI

1. Subscription and Tool Consolidation

Identify opportunities to replace multiple existing subscriptions with a single, comprehensive AI tool. This includes consolidating tools for tasks like note-taking, business intelligence, data processing, and even image sourcing (e.g., stock image libraries). This is your easiest ROI win—it's purely financial and immediately measurable.

2. Incremental Task Efficiency

Target improvements in repetitive tasks by aiming for a modest 20% efficiency gain per department or team. For a task that takes an hour daily, a 15-minute time savings is a quantifiable success. This compounds quickly: a single administrative assistant saving 15 minutes daily represents approximately 60 hours per year, or roughly 1.5 weeks of working time.

3. Quality and Accuracy Enhancement

Leverage AI to double-check and validate human output, particularly in tasks involving visual analysis. Implement mobile AI tools for functions like advanced spell and grammar checking from a photo, or build AI agents to analyze incoming documents (such as checks or invoices) and automatically export the reconciled data to a spreadsheet or database. Quality improvements directly reduce errors, rework, and downstream costs.

Fostering a Culture of Measurable Success

Once the Return on Investment (ROI) of an AI solution is demonstrated, empower your team by granting them the freedom to experiment with the time that has been saved. Your next AI innovations and "heroes" may emerge from the most unexpected places.

A key principle of your company culture should be the commitment to measuring every success. Specifically, reward employees who develop solutions that solve a problem and deliver demonstrable ROI. By incentivizing measurable results, you will motivate your people to actively seek them out.

A Real Example from Our Own Team

Consider the example from our own team: the production of our weekly "AI Insights" blog. The time commitment for this task was reduced from approximately 8 hours per week to just a few hours. While we dedicate the same amount of time to the intellectual work of writing, we now leverage AI for drafting, styling, publishing the post, and emailing subscribers. This efficiency gain was quantifiable precisely because we are a culture of measurers! We established a baseline by timing the initial manual process (using a time-tracking application) and now measure the time saved through automation.

That's approximately 20 hours saved per month, or 240 hours per year. At an average salary of just $60,000 per year, that's roughly $7,200 in value delivered annually—from a single workflow optimization.

Conclusion

Measuring ROI in AI isn't just about proving the value of technology—it's about cultivating a culture of measurement, alignment, and continuous improvement. By setting clear objectives, appointing the right leaders for both strategic and project-level initiatives, and focusing on small, measurable wins, organizations can demystify AI's impact.

Over time, these wins compound, building confidence in AI investments and empowering teams to innovate responsibly. The ultimate goal is not only to achieve measurable ROI but to embed a mindset where every success is tracked, celebrated, and used as a stepping stone toward greater organizational excellence.