A guide for Irish business leaders navigating the tension between AI ambition and tangible returns

There’s a peculiar contradiction happening in boardrooms across Ireland right now. Nearly three-quarters of organisations report that their most advanced AI initiatives are meeting or exceeding ROI expectations. Yet, when you scratch beneath the surface, most leadership teams are simultaneously paralysed by a seemingly simple question:

What should we actually do with AI?

If you’re still asking this question in 2025, you’re already behind. Your competitors aren’t asking. They’re executing. And while you’re scheduling another exploratory meeting, they’re capturing market share with AI capabilities you’re still “evaluating.”

This isn’t a failure of imagination. Irish businesses are brimming with ideas like automating customer service, personalising marketing, optimising supply chains, and accelerating product development. The problem is more fundamental and more damning: you’re treating AI as a technology problem when it’s actually a strategy problem. And you’re avoiding the strategy problem because it requires making difficult choices you’re not prepared to make.

Recent research from Deloitte’s AI Institute reveals that the organisations successfully extracting value from AI aren’t necessarily the ones with the most sophisticated algorithms or the deepest technology budgets. They’re the ones with something far more valuable: strategic clarity about what AI should accomplish and disciplined orchestration of competing priorities.

In other words, they’re thinking like Chief Strategy Officers, whether they have one or not.

The Four Tensions That Are Quietly Sabotaging Your AI Ambitions

If your organisation is struggling to move AI initiatives from pilot to production, from experiment to enterprise value, you’re likely caught in one or more of four strategic tensions. And here’s the uncomfortable truth. These tensions don’t resolve themselves. They require active, strategic decision-making.

Tension #1: The Portfolio Paradox—Build Everything or Focus on Anything?

Walk into any mid-to-large Irish enterprise today and you’ll find AI initiatives sprouting like mushrooms after rain. Marketing has a chatbot pilot. Operations is experimenting with predictive maintenance. HR is testing resume screening tools. Finance is exploring fraud detection. IT is evaluating infrastructure options.

Each initiative has merit. Each has a business case. And collectively, they’re creating chaos.

The portfolio paradox is this: innovation requires experimentation, but value requires focus. You need both, but the balance point isn’t apparent, and it shifts constantly based on your market position, competitive threats, and organisational maturity.

Consider the contrasting approaches revealed in the Deloitte research. One media and technology enterprise operates with surgical precision: “A top-down enterprise strategy guides everything. We articulated investment areas in each of our five growth markets.” Their AI portfolio isn’t a collection of hopeful experiments; it’s an integrated component of capital planning and M&A strategy.

Contrast this with a consumer brands company taking a hybrid approach: “We have a top-down strategy of areas where we can strengthen with AI, but there is grassroots innovation happening. People are exploring use cases within their own functions.” They’ve created governance forums where innovation can surface organically while maintaining strategic guardrails.

Neither approach is inherently superior. The question is: which approach serves your strategic objectives?

Here’s the challenge most Irish leaders aren’t addressing: Is experimentation valuable in its own right, or should every initiative be tied to near-term ROI? That’s a strategic choice, not a financial calculation. If you’re in a market facing imminent disruption from AI-native competitors, you may not have the luxury of patience. If you’re in a stable market where AI offers an incremental advantage, premature focus might actually stifle the experimentation that drives breakthrough applications.

The emarkable question: Do you have a strategic framework for evaluating which AI initiatives deserve investment, or are you defaulting to whoever makes the most compelling presentation? If it’s the latter, you’re not building an AI strategy. You’re building a portfolio of hopes.

Tension #2: The ROI Riddle—When the Numbers Keep Changing

Let’s address the elephant in every AI strategy meeting: nobody really knows what this is going to cost.

You can pilot an AI application, measure productivity gains, and build a business case based on extrapolation. But then the pricing model changes. Or the technology vendor pivots. Or a new model emerges that’s 10x more efficient at half the cost. Or regulation shifts the compliance burden. Or your competitors deploy a similar capability, eliminating your differentiation.

One CSO captured this frustration perfectly:

“Part of the reason this is so hard is that I can make all the assumptions in the world about what I believe the benefit to be, but the difficulty is not in value creation. It is in how the AI ecosystem is evolving. Large ecosystem players haven’t figured out yet how to price any of this.”

This creates a strategic dilemma: Do you move fast and accept uncertainty, or do you wait for clarity and risk being left behind?

The uncomfortable answer is that different initiatives require different approaches. Some AI applications, particularly those focused on operational efficiency with clear metrics, can be evaluated with traditional ROI frameworks. If an AI tool demonstrably reduces customer service response time by 40% and you can calculate the cost savings, the decision is relatively straightforward.

But what about AI initiatives aimed at competitive differentiation? How do you calculate the ROI of not losing market share to an AI-native competitor? How do you quantify the value of learning and organisational capability building?

Here’s what separates strategic leaders from tactical executors: they understand that not all AI investments are created equal, and they shouldn’t all be evaluated with the same ROI lens.

Some investments are about survival, responding to competitive threats or meeting emerging customer expectations. Some are about positioning, building capabilities that create optionality for future moves. Some are about efficiency, extracting cost from existing operations. Each requires a different risk tolerance, time horizon, and success metrics.

The research reveals that strategy development timelines are shrinking for 68% of organisations, with half experiencing increased strategy refresh frequency. This isn’t academic; it means your AI investment decisions need to account for accelerating change. A three-year payback period might sound conservative, but if the competitive landscape shifts in 18 months, conservative becomes reckless.

The emarkable challenge: Are you making AI investment decisions based on static ROI calculations, or do you have a dynamic framework that accounts for strategic value, competitive positioning, and organisational learning? If you’re still using the same financial models you used for ERP implementations, you’re optimising for the wrong outcomes.

Tension #3: The Governance Gamble—Innovation or Insurance?

Every organisation says they want to “move fast and break things” with AI right up until something breaks that matters, customer trust, regulatory compliance, data security, and brand reputation.

The tension between innovation velocity and risk management is real, and it’s creating decision paralysis in Irish boardrooms. Some leaders adopt a “build first, govern later” mentality, believing that overcoming inertia requires accepting some risk. Others implement governance frameworks so comprehensive that innovation suffocates under the weight of compliance.

Consider these contrasting philosophies from actual CSOs:

One technology leader described navigating constant strategic tradeoffs:

“If we decide with a product that we won’t retain any customer data to provide maximum security, the flip side is that it reduces our ability to develop the product and create cross-customer benefits because the system cannot learn from multiple customers.”

Meanwhile, a news media executive takes an explicitly value-first approach:

“What I don’t want ever to do is have a self-censoring conversation about something that sounds promising. If there is a promising use case, I don’t want to talk about theoretical problems. Find the shortest path from ‘here’s an idea’ to ‘here’s how.'”

Both can’t be right. Or can they?

The strategic insight here is that governance isn’t a universal standard; it’s a context-dependent choice that should reflect your risk tolerance, industry requirements, and competitive position.

If you’re a financial services firm, the regulatory burden and reputational risks of AI failure probably necessitate a governance-first approach. If you’re a media company competing with AI-native content creators, innovation-first might be existential.

But here’s what’s non-negotiable: you need to make this choice explicitly and strategically. Most organisations default to whatever their most risk-averse executive advocate for, or whatever their most innovation-hungry leader can sneak past compliance. Neither approach is strategic.

The emarkable reality check: Do you have an explicit governance philosophy for AI that reflects your strategic priorities, or are you reactively implementing policies in response to whoever raised the loudest concern in the last meeting? If your answer involves the phrase “better safe than sorry,” you’ve already chosen a side, and it might be the wrong one for your competitive context.

Tension #4: The Talent Transformation—Everyone Needs to Change (But Nobody Wants To)

And yet, most AI strategies treat workforce transformation as an afterthought, something HR can handle with a few training sessions and some change management communications.

This is strategic malpractice.

The research reveals that 54% of CEOs now rank talent acquisition and transformation in their top three priority areas, on par with technology investment itself. This isn’t because leaders suddenly developed a compassionate interest in employee development. It’s because they’re realising that the technology is the easy part; the transformation is the hard part.

You can deploy the most sophisticated AI infrastructure in your industry. Still, if your workforce doesn’t know how to leverage it, doesn’t trust it, or actively resists it, you’ve built an expensive monument to wasted potential.

One technology CSO captured the scope of this challenge: “The amount of change they are having to embrace and deal with is energising for some, but no matter what, 100% of the employee base will have to use AI in daily work. Everyone needs to upskill, and these things are new for everyone.”

The strategic question isn’t whether to invest in workforce transformation; that’s table stakes. The question is how to sequence these investments relative to technology deployment and accelerate adoption without breaking organisational culture.

Some organisations are taking a “show, don’t tell” approach, deploying AI tools that demonstrably improve daily work and using early success stories to build momentum. Others are investing heavily in upfront training and change management, believing that capability building must precede deployment.

Again, neither approach is universally correct. The right answer depends on your organisational culture, change readiness, and strategic timeline.

But here’s what the successful organisations understand: workforce transformation isn’t a one-time programme, it’s an ongoing strategic priority that requires the same rigour and investment as the technology itself.

The emarkable question: When you calculate the ROI of AI initiatives, are you including the full cost and timeline of workforce transformation? Or are you systematically underestimating the difficulty of change and wondering why adoption falls short of projections?

Five Strategic Principles for AI Success (That Most Irish Leaders Are Ignoring)

These tensions are supposed to be painful. They’re strategic dilemmas without clean answers, which means they require judgment, trade-offs, and ongoing adjustment.

But while there’s no universal playbook, there are strategic principles that separate organisations extracting value from AI from those drowning in pilot purgatory:

1. Strategic Objectives Drive AI Decisions (Not the Other Way Around)

This seems obvious, yet most organisations constantly violate this principle. They pursue AI initiatives because the technology is cool, or because competitors are doing it, or because vendors are selling it persuasively.

AI should be an enabler of enterprise strategy, a catalyst for rethinking business models, value propositions, and operational capabilities. Every AI initiative should answer the question: “How does this advance our strategic objectives?”

If you can’t articulate that connection clearly, you’re experimenting without a strategy.

2. Think “String of Pearls”, Not Monolithic Transformation

Stop thinking about “AI strategy” as a singular thing. Instead, think of it as a portfolio of distinct initiatives, a “string of pearls”, each with its own business case, ROI, risk profile, and timeline.

Some pearls are about efficiency. Some are about differentiation. Some are about learning and capability building. Evaluate each on its own merits, but ensure the collective string adds up to meaningful enterprise value.

3. The CSO as Scorekeeper (Even If That’s Not Your Title)

Someone in your organisation needs to play the role of scorekeeper, constantly connecting strategic objectives to execution activities and holding leaders accountable for outcomes.

Is upskilling progressing as expected? Are pilots languishing without deployment decisions? Are deployed use cases delivering projected value? These aren’t comfortable questions, but they’re essential.

If you don’t have a CSO, this role might fall to you. And if it falls to nobody, your AI strategy will drift aimlessly between competing priorities.

4. Weigh the Risk of Disruption (Not Just the Promise of Opportunity)

Most AI business cases focus on upside opportunity, revenue growth, cost reduction, and productivity gains. But for many Irish businesses, the more critical calculation is downside risk.

If competitors deploy AI capabilities that fundamentally change customer expectations or cost structures in your industry, the risk of inaction might far exceed the risk of imperfect action.

This requires honest assessment: Is your industry facing imminent AI disruption? If so, your risk tolerance and urgency should reflect that reality.

5. Risk and Compliance Are Strategic Choices (Not Bureaucratic Burdens)

Every AI initiative carries risk to data security, customer trust, regulatory compliance, and operational continuity. The question isn’t whether to manage these risks; it’s when and to what extent.

Some risks need to be addressed up front. Others can be managed iteratively as you scale. The decision about which is which is a strategic choice that should reflect your competitive context and risk tolerance, not a default to maximum caution or reckless speed.

The Strategic Assistance Gap: Why You Can’t Do This Alone

Here’s the final uncomfortable truth: Most Irish organisations don’t have the internal strategic capacity to navigate these tensions effectively.

This isn’t a criticism. It’s a recognition of reality. Strategic AI orchestration requires simultaneously understanding:

• Technology capabilities and limitations
• Business model implications and competitive dynamics
• Financial modelling under uncertainty
• Change management and organisational psychology
• Risk management and regulatory compliance
• Vendor ecosystem evolution and market dynamics

No single executive possesses all this expertise. Even organisations with dedicated strategy teams often lack the specific AI domain knowledge to make informed trade-offs.

This is precisely where external strategic assistance becomes not just valuable, but essential.

The right strategic partner doesn’t just implement AI solutions; it also helps drive innovation. They help you navigate the tensions between competing priorities. They help you:

• Develop strategic frameworks for evaluating and prioritising AI initiatives based on your specific competitive context
• Model scenarios and ROI under different assumptions about technology evolution and market dynamics
• Design governance approaches that reflect your risk tolerance and strategic ambitions rather than generic best practices
• Accelerate workforce transformation with proven change management approaches tailored to your organisational culture
• Monitor the competitive landscape to identify emerging threats and opportunities before they become existential

Most importantly, they bring pattern recognition from working across industries and organisations. Perspective you can’t develop in isolation.

Your Strategic Call to Action: Three Questions You Need to Answer This Quarter

Let me leave you with three questions that will reveal whether you’re genuinely thinking strategically about AI or just reacting to pressure:

Question 1: Can you articulate, in two sentences, how AI advances your three most important strategic objectives over the next 18 months?

If you can’t answer this clearly, you don’t have an AI strategy; you have a collection of initiatives.

Question 2: Do you have explicit criteria for deciding which AI initiatives to scale, which to kill, and which to let continue experimenting?

If every initiative remains in permanent pilot status, you’re avoiding strategic choices, not making them.

Question 3: Who in your organisation is responsible for orchestrating AI initiatives across competing priorities and holding leaders accountable to strategic outcomes?

If the answer is “everyone” or “nobody in particular,” you’ve identified your most enormous strategic gap.

The emarkable Invitation: Let’s Get Strategic Together

At emarkable, we don’t believe in AI for AI’s sake. We believe in strategic clarity that drives business outcomes. We think that achieving that clarity requires the courage to confront uncomfortable tensions rather than avoid them.

If reading this piece has left you:

• Questioning whether your AI initiatives are truly strategic
• Recognising tensions in your approach that nobody is actively managing
• Wondering whether you’re making trade-offs by default rather than by design
• Concerned that your organisation lacks the internal capacity to navigate these complexities

Then let’s talk.

We’re not interested in selling you technology implementations you don’t need. We’re interested in helping Irish business leaders develop the strategic clarity and orchestration capability to extract genuine value from AI investments.

Because here’s the truth that nobody wants to say out loud: Most AI initiatives fail not because of technology limitations, but because of strategic incoherence.

The question isn’t whether AI will transform your industry; it will. The question is whether you’ll shape that transformation strategically or react to it desperately.

The organisations that emerge as leaders in the AI era won’t be the ones with the most significant AI budgets or the most impressive pilots. They’ll be the ones with the strategic clarity to make hard choices, the orchestration capability to align competing priorities, and the humility to seek assistance when internal capacity falls short.

Which type of organisation will yours be?

Ready to move from AI ambition to AI strategy? Contact emarkable to schedule a strategic AI assessment. We’ll help you identify the tensions holding you back and develop a framework for making trade-offs that advance your most important objectives.

Because strategy isn’t about having all the answers, it’s about asking the right questions, and having the courage to act on what they reveal.

---

emarkable.ie – Strategic AI Partnership for Irish Business Leaders