Rethinking Business Cases in the Age of AI: Finding High-Value AI Opportunities

In my previous articles on AI business cases, I explored why traditional approaches fall short and outlined the essential building blocks for more effective evaluation. While the approach helps when assessing AI initiatives more comprehensively, it raises an equally important question: how do you find the right opportunities to evaluate in the first place? Without a systematic approach to identifying high-potential AI initiatives, organisations risk applying even the best business case frameworks to the wrong opportunities.
I’ve observed this challenge often in my work with multiple businesses across industries. The allure of highly visible AI applications often overshadows deeper strategic opportunities. Chatbots provide a perfect example of this phenomenon - while they deliver quick returns through enhanced customer service efficiency, their prominence, a trend identified in 2023 that persists today amid ongoing generative AI enthusiasm, often distracts from high-impact use cases like process automation, supply chain optimisation, or advanced analytics, which offer greater long-term value despite requiring more complex integration.
Organisations frequently rush to deploy chatbots without considering whether other AI investments might deliver substantially greater value. A predictive maintenance system might require more upfront investment and technical complexity, but could transform operational reliability, extend equipment lifespans, and enable entirely new service-based revenue models. Similarly, AI-enhanced pricing optimisation might lack the visible showcase appeal of a chatbot but could drive margin improvements that dwarf the cost savings from automated customer interactions. The issue isn’t that these more visible applications lack value, but that their prominence in the AI conversation can distract from more transformative opportunities that better align with strategic priorities.
To avoid this trap, organisations need a systematic approach to identify AI opportunities that deliver transformative value, starting with a rigorous evaluation of their business processes.
Process Evaluation and Analysis
At its core, identifying AI opportunities begins with understanding your processes. The most successful organisations start by examining their operations through a structured lens, looking for characteristics that indicate high potential for AI enhancement.
When identifying candidate opportunities for AI, I centre my evaluation on five critical dimensions which help me pinpoint areas where AI can create disproportionate value:
- Decision complexity: Processes involving nuanced, multi-factor decision-making, particularly where expert judgment navigates shades of grey rather than binary choices often benefit significantly from AI. Insurance underwriting exemplifies this complexity, with underwriters evaluating dozens of variables that interact in subtle ways. AI can enhance these decisions by detecting patterns humans might miss while still allowing expert judgment for edge cases.
- Information volume: Processes drowning in data; from customer interactions to equipment telemetry, frequently struggle with human-scale analysis limitations. Consider airline flight operations, where teams must simultaneously monitor thousands of data points across aircraft performance, weather patterns, passenger connections, crew scheduling, and airport constraints. AI’s ability to process this information at scale creates substantial value by identifying subtle patterns and anomalies that human operators might miss. For instance, AI systems can predict cascading delay effects across a network hours before they materialise, allowing for proactive rescheduling that preserves operational integrity and customer experience in ways that would be impossible with manual analysis alone.
- Process variability: Highly variable processes – those requiring case-by-case adaptation rather than standardised approaches – often create significant cognitive load for workers. Healthcare diagnosis exemplifies this variability, with each patient presenting unique symptom combinations and medical histories. AI can support clinicians by suggesting potential diagnoses based on similar historical cases while adapting to each patient’s specific context.
- Cognitive repetition: Processes requiring humans to perform the same analytical operations repeatedly, for example, document review, report generation, or requirements scanning, represent prime AI enhancement opportunities. Legal contract review illustrates this pattern, with associates spending countless hours performing essentially the same cognitive task across different documents. AI can accelerate these tasks dramatically while freeing human experts to focus on more creative or complex aspects of their work.
- Knowledge continuity risk: Processes where critical knowledge resides in individuals’ heads rather than documented systems create significant organisational vulnerability. Technical support for legacy equipment often faces this challenge, with a shrinking pool of experts who understand historical systems. AI can codify this expertise, making it accessible to newer staff while preserving continuity as experienced workers retire.
By systematically evaluating your processes through these dimensions, you create a structured opportunity identification approach. I typically recommend using a simple scoring system for each dimension (high/medium/low) and focusing first on processes that score highly across multiple dimensions. This evaluation should involve cross-functional teams with both operational and technical perspectives to ensure comprehensive assessment.
For large enterprises, this evaluation may need to consider thousands of business processes across multiple divisions and geographies. The key is developing a tiered approach that begins with broad process categories before drilling down into specific sub-processes with the highest potential. This prevents the assessment from becoming overwhelming while ensuring no high-value opportunities are overlooked.
This approach can be particularly revealing in complex industries. For instance, imagine how an airline might evaluate its key operational processes using this framework. While the executive team might initially focus on customer-facing applications like chatbots for handling routine inquiries, a systematic analysis could reveal that areas like aircraft maintenance and crew scheduling present significantly higher AI value potential.
Consider how five illustrative airline processes might score when evaluated through this framework:
Process | Complexity | Volume | Variability | Cognition | Knowledge | Score |
---|---|---|---|---|---|---|
Crew Scheduling | High | High | High | Medium | Medium | 13/15 |
Aircraft Maintenance | High | High | Medium | High | High | 14/15 |
Revenue Management | High | High | Medium | Medium | Low | 11/15 |
Passenger Service | Medium | Low | High | High | Low | 10/15 |
Ground Operations | Medium | Medium | High | Medium | Medium | 11/15 |
In this example, aircraft maintenance emerges as the highest-potential AI opportunity, scoring high on almost every dimension. The process involves complex decisions about when and how to perform maintenance, processes massive volumes of sensor and performance data, requires specialists interpreting diverse signals, and faces significant knowledge continuity challenges as experienced engineers retire. From this process evaluation, we can define a specific AI initiative concept: “Predictive Aircraft Maintenance AI.” which could potentially transform reliability, reducing unscheduled maintenance events while extending aircraft service life.
This objective evaluation helps overcome organisational biases that might otherwise direct AI investments toward the most visible or politically advantageous areas rather than those with the greatest potential value. It also provides a transparent rationale for investment decisions that can be communicated across the organisation, building understanding and alignment behind the selected opportunities.
Strategic Alignment Assessment
Once we’ve identified high-potential processes like aircraft maintenance, the next step is evaluating how specific AI initiatives derived from these processes align with broader organisational objectives. This is where the Well-Advised Framework proves particularly valuable for opportunity screening.
Using this framework, boards can evaluate potential AI initiatives across five key pillars: Innovation and New Products/Services, Customer Value and Growth, Operational Excellence and Efficiency, Responsible Business Transformation, and Revenue, Margin, and Profit. The most valuable AI opportunities aren’t those that score highest in a single pillar but those that deliver meaningful impact across multiple pillars.
Taking our “Predictive Aircraft Maintenance AI” initiative as an example, let’s examine how it aligns across the five strategic pillars:
Well-Advised Pillar | Predictive Maintenance AI Value | Potential Impact | Score (1-5) |
---|---|---|---|
Innovation and New Products/Services | Potential to develop new service offerings around guaranteed aircraft availability and maintenance-as-a-service for other airlines | New revenue streams, industry leadership position | 3/5 |
Customer Value and Growth | Improved on-time performance, decreased cancellations, enhanced safety record | Higher customer satisfaction, reduced compensation costs, reputation enhancement | 4/5 |
Operational Excellence and Efficiency | Reduction in unscheduled maintenance events, optimised maintenance scheduling, extended aircraft lifespan | Lower maintenance costs, increased aircraft availability, improved asset utilisation | 5/5 |
Responsible Business Transformation | Reduced environmental impact through more efficient operations and fewer ferry flights for maintenance | Sustainability improvements, regulatory compliance, reduced carbon footprint | 4/5 |
Revenue, Margin, and Profit | Operational cost savings, additional revenue from higher aircraft availability, potential new service revenue | Improved margins, cost avoidance, multiple revenue impact points | 4/5 |
I use a simple scoring matrix to visualise this alignment. For each potential AI opportunity, I assess its potential contribution to each pillar on a 1-5 scale. The resulting pattern reveals not just the magnitude of potential impact but its strategic balance. A balanced initiative that scores 3-4 across all pillars often creates more meaningful value than one scoring 5 in a single dimension but 1-2 in others.
Overall Strategic Alignment Score: 20/25
This strategic alignment assessment reveals that what might initially appear to be a purely operational initiative actually delivers substantial value across all five pillars. Without this structured evaluation, the airline might have underestimated the initiative’s strategic importance, potentially viewing it merely as a maintenance cost reduction tool rather than a multi-dimensional strategic asset.
The assessment also helps identify potential enhancements to the initiative. For example, the relatively lower score in Innovation suggests an opportunity to explore how the predictive maintenance capability could be packaged as a service for other airlines or how the underlying algorithms might be applied to other operational areas.
This approach helps Boards avoid the common trap of over-investing in cost-reduction focused AI while neglecting opportunities that drive innovation or customer value. It also surfaces those rare opportunities that deliver across multiple strategic dimensions – the true transformative initiatives that deserve priority investment - when viewed through a broader lens.
Without a structured alignment assessment that systematically examines potential across all dimensions, organisations routinely miss these valuable connections that transform single-purpose tools into multidimensional strategic assets. The table illustrates how what might initially appear to be a narrow operational improvement actually delivers strategic value across every dimension of the business, with particularly strong contributions to operational excellence, customer value, and financial performance.
Implementation Feasibility Analysis
Strategic value means little without implementation feasibility. The third critical lens for opportunity screening examines execution risk through the Five Pillars of AI capability: Governance & Accountability, Technical Infrastructure, Operational Excellence, Value Realisation & Lifecycle Management, and People, Culture, & Adoption.
Continuing with our Predictive Aircraft Maintenance AI example, let’s examine how an airline might assess its readiness to implement this strategically valuable initiative:
Capability Pillar | Current Readiness | Key Gaps | Feasibility Score |
---|---|---|---|
Governance & Accountability | Medium | Lack of clear governance framework for AI-driven maintenance decisions; undefined human oversight protocols for safety-critical systems | 3/5 |
Technical Infrastructure | High | Strong existing sensor networks on aircraft; robust data collection systems; established cloud infrastructure | 4/5 |
Operational Excellence | Medium | Limited MLOps practices; need for integration with existing maintenance planning systems | 3/5 |
Value Realisation & Lifecycle Management | Low | No established methodology for measuring AI impact on maintenance outcomes; unclear model monitoring approach | 2/5 |
People, Culture, & Adoption | Medium | Maintenance engineers have technical expertise but limited AI familiarity; potential resistance to changing established safety protocols | 3/5 |
Overall Feasibility Score: 15/25
This assessment reveals important insights that might not be captured in traditional business case evaluations. While the Predictive Aircraft Maintenance AI initiative offers exceptional strategic value (20/25), its implementation feasibility is more moderate (15/25). However, this doesn’t mean the initiative should be abandoned. Rather, it indicates specific capability gaps that need addressing for successful implementation.
For example, the airline has strong technical infrastructure in place - a crucial foundation - but needs to develop better governance frameworks and value measurement systems. These capabilities would need to be built alongside the technical solution, not just as afterthoughts.
The feasibility analysis also helps sequence development work. In this case, the airline might start by establishing an AI Centre of Excellence (AI CoE) reporting directly to the board’s risk committee. This AI CoE would develop clear governance protocols for AI maintenance applications and design value measurement frameworks before fully scaling the solution. By establishing this foundational structure first, the airline ensures coordinated capability development across all five pillars rather than focusing solely on the technical implementation. This sequencing ensures they build the necessary capabilities to succeed before major investments, while creating reusable governance patterns for future AI initiatives.
This approach aligns with the airline’s current position in the AI Stages of Adoption. If they’re in the early Adopting stage, this initiative provides an opportunity to develop crucial capabilities around AI governance and value measurement that will benefit future initiatives beyond maintenance. By addressing these capability gaps, they can move more confidently toward the Optimising stage.
The feasibility lens ensures that boards don’t just chase the theoretically highest-value opportunities but focus on those most likely to succeed given their organisation’s current capability profile. The ideal AI initiatives are those that deliver value while simultaneously building capabilities that enable future innovations.
Portfolio Approach to AI Initiatives
Rather than evaluating AI opportunities in isolation, successful organisations build balanced portfolios. This approach recognises that different classes of AI initiatives serve different strategic purposes and should be managed accordingly.
I typically recommend categorising opportunities into three broad groups:
The first category comprises quick wins focused on efficiency and cost reduction. These initiatives typically deliver measurable financial returns within months, involve limited organisational change, and face minimal technical complexity. They build momentum and credibility for your AI programme while generating resources for more ambitious efforts. Customer service chatbots or document automation applications often fall into this category. While their individual impact may be modest, collectively they create savings that can fund more transformative initiatives.
The second category focuses on capability building and medium-term value. These initiatives might not deliver immediate financial returns but create the foundational capabilities required for longer-term transformation. Data architecture modernisation, governance framework development, or AI skills programmes exemplify this category. Their value comes from enabling future opportunities rather than direct financial impact, making them difficult to justify through traditional business cases.
The third category encompasses transformative initiatives with longer horizons. These opportunities reimagine core business processes, enable new business models, or create sustainable competitive advantage. Their implementation complexity and time horizons exceed most traditional investment frameworks, but their potential impact justifies the investment. AI-driven personalisation that fundamentally transforms customer relationships or predictive maintenance systems that enable new service-based business models exemplify this category.
A balanced AI portfolio should include initiatives from all three categories. While the precise distribution depends on your organisation’s strategic context and AI maturity, I typically recommend that organisations in the Experimenting or Adopting stages allocate 60-70% to quick wins, 20-30% to capability building, and 10-15% to transformative initiatives. As organisations progress to Optimising and Transforming stages, this distribution shifts toward more ambitious opportunities.
This portfolio approach ensures continuous delivery of visible value while building toward more significant transformation. It provides boards with both quick wins to demonstrate value and strategic initiatives to create sustainable advantage.
Putting It All Together: A Systematic Identification Process
The approach set out here creates a systematic process for identifying high-value AI opportunities. Rather than relying on random suggestions or hype-driven hunches, this approach ensures comprehensive exploration of potential value creation areas.
Start with process analysis to identify areas with high enhancement potential based on decision complexity, information volume, process variability, repetitive cognitive tasks, and knowledge continuity risk. This analysis creates your initial opportunity pool.
Next, apply strategic alignment analysis using the Well-Advised framework to prioritise opportunities based on their contribution to your organisation’s strategic objectives. Focus on opportunities that deliver across multiple pillars rather than just those with the highest single-dimension score.
Then, conduct feasibility analysis using the Five Pillars framework to understand implementation risks and prerequisites. This helps sequence opportunities based on both value potential and execution feasibility.
Finally, build a balanced portfolio that delivers immediate value through quick wins while investing in capability building and transformative initiatives. This ensures continuous delivery while working toward longer-term objectives.
Implementing this structured approach will help you identify significantly higher-value opportunities than relying on ad hoc methods.
Conclusion: From Opportunity Identification to Business Case Development
Finding the right AI opportunities lays the essential foundation for effective business case development. By applying a systematic identification process that examines processes, strategic alignment, implementation feasibility, and portfolio balance, boards can ensure they’re evaluating the opportunities with the greatest potential value and execution likelihood.
This approach transforms AI business case development from a reactive process – evaluating proposals as they emerge – to a proactive one that systematically surfaces and prioritises high-potential opportunities. It ensures that your carefully crafted business case frameworks are applied to the initiatives most likely to create sustainable value.
In the next article in this series, I’ll provide a step-by-step guide for building comprehensive AI business cases that boards can confidently evaluate. We’ll explore how to structure these business cases to address the unique characteristics of AI investments, incorporate the evaluation frameworks we’ve discussed, and create compelling narratives that drive informed decision-making. This practical guide will help you translate opportunity identification into actionable investment proposals that balance financial rigour with strategic vision.
Let's Continue the Conversation
I'm interested in hearing about your organisation's approach to identifying AI opportunities. What methods have you found most effective? Where do you find the greatest challenges in prioritising potential AI initiatives?
About the Author
Mario Thomas is a transformational business leader with nearly three decades of experience driving operational excellence and revenue growth across global enterprises. As Head of Global Training and Press Spokesperson at Amazon Web Services (AWS), he leads worldwide enablement delivery and operations for one of technology's largest sales forces during a pivotal era of AI innovation. A Chartered Director and Fellow of the Institute of Directors, and an alumnus of the London School of Economics, Mario partners with Boards and C-suite leaders to deliver measurable business outcomes through strategic transformation. His frameworks and methodologies have generated over two-billion dollars in enterprise value through the effective adoption of AI, data, and cloud technologies.