Why So Many Smart Companies Are Failing At AI Deployment & How To Get AI-Powered Innovation & Digital Transformation Right

What Causes Poor ROI from AI Investments & failed digital transformations?

As an AI keynote speaker and an expert on technology-powered innovation and digital transformation, I am witnessing companies investing vast sums in artificial intelligence pilots, infrastructure, and tools, yet not seeing meaningful returns. According to an MIT study that made big headlines, 95% of AI pilot projects are failing to deliver positive ROI.

As I detail in a companion piece to this article, on what I call The LEADERSHIP-AI Synthesis, many AI experiments are actually costing businesses time and trust due to errors, slop, and cognitive decay rather than delivering a great ROI. This is not because AI lacks potential. It’s because most organizations are trying to use 21st-century technology to run 20th-century business and operating models.

This is not the first time I have seen the desire to quickly pick up a new and disruptive technology without really understanding how it can add benefits to the value proposition of a business or internal function. The same thing happened with Web 1.0, Web 2.0, SMS, mobile computing, apps, and Web 3.0, too.

I've seen these strategic shortcuts—attempting to get a massive ROI from significant investments in tech but not enough in strategic thinking—arise time and time again when Fortune 500s and FTSE 250s come to my innovation consultancy with the latest tech buzz front of mind and ask us to leverage it to create outsized value... without going on the journey that all successful innovation and digital transformation require.

Simply put, you can’t just shove AI—or any emerging or exponential technology—into an existing workflow or culture and hope it’s going to unlock a 10x ROI. To unlock real and sustained value, you have to redesign how your business works—and rethink what business you are really in or should be in—around potentially transformative technologies.

Those who take the time to do deep strategic thinking, customer-centric creativity, and systemic transformation work—and have built up deep and sustainable capabilities in innovation and transformational leadership—can drive disruptive ideas and digital transformations forward and reap the rewards of 10x and even 100x returns. Those that do not, by and large, cannot.

Think about digital-first banks—such as Revolut or Wise—which can open a corporate account in minutes through their AI-powered digital security and compliance innovations. Legacy banks, even the behemoths, still ask business customers to come into a store to sign documents or send forms by mail.

For years, I have heard the same refrain from the senior execs in banking I worked with on innovation leadership and digital transformation: "Our industry is different, it's highly regulated, we cannot innovate like other sectors, our IT systems cannot do X, etc." Meanwhile, more nimble and innovative competitors outsmarted them.

It is this lack of strategic vision (often, in truth, complemented by a touch of arrogance, avoidance, and conceptual obsolescence) that underlies so many AI investment failures within organizations. Execs want quick, silver-bullet fixes that reduce costs and increase margins... rather than leading the journey of becoming a transformed business that shapes the future of an industry.

This is a big problem for businesses and an even bigger problem for the global economy, as growth is becoming increasingly reliant on AI investment returns. Bain & Company’s Global Technology Report estimates that the AI industry must generate $2 trillion in new annual revenue by 2030 to cover its infrastructure and compute investments, five times the size of today’s global subscription-software market. Even then, Bain predicts an $800 billion shortfall between projected efficiency savings and the capital required to sustain AI growth.

From decades of consulting to large corporations seeking to capitalize on the productivity and performance promises of new technologies, I have learned that retrofitting AI into existing workflows is unlikely to deliver exponential value. This is just one of the 5 Ways Corporate Innovation Activities Are Likely To Fail.

To unlock genuine and sustained ROI from AI investments, leaders must redesign their business, operating, and people models around what AI—whether predictive, generative, or agentic AI—and other technologies can really enable.

The Mindsets Needed for AI-Driven Digital Transformation & Innovation

Most AI initiatives fail because organizations treat AI as a plug-and-play productivity tool rather than a catalyst for transformation. They hope AI will optimize existing processes (and reduce headcount and costs) rather than reimagine how value is created.

Real AI success through breakthrough innovation and business transformation demands a shift from mechanistic efficiency thinking to creative systems thinking, re-architecting the business (and this also means its culture) rather than inserting a new piece of tech into the old machinery and hoping for the best.

In the history of technology, outdated mindsets are always what hold back innovation growth. For example, even after the printing press democratized access to books across Europe, monks continued to illustrate books by hand, slowing adoption, production, and distribution. Cultural habits, mental models, and the desire to protect their status slowed the technology-driven transformation.

Beyond obvious use cases such as logistics, overhead reductions, and supply chain efficiencies—which are not always that obvious given the error rates and workflow integration challenges of AI—businesses must start not with the technology itself, but with the humans who will use it to make, buy, sell, and service your products and services.

If the humans do not think an AI tool or AI-powered process helps them to solve problems they cannot solve themselves, they will not adopt it. They may even reject it.

In the AI leadership programs and digital transformation/innovation processes I co-design and co-lead, rather than rushing to start a pilot, we slow executives and technologists down and ask questions like:

• Why do we want to use AI beyond personal effectiveness?

• How do customers, users, and employees currently engage with your business?

• What are their pain points and unmet needs that AI could address?

• What are emerging pain points in new customer segments that AI could solve?

• What could an entirely new customer or employee experience look like if predictive AI, automation, generative AI, and agentic AI were designed from the ground up?

• What would the value of these be against existing user experiences?

Only then can AI become a true force for growth, innovation, and transformation rather than an expensive and often dangerous investment that will need to be written off.

Why a good roi from AI Requires Real Transformation, Not Quick Fixes

I’ve been designing and leading technology-powered innovation and digital transformation programs for more than 25 years, helping organizations leverage successive waves of tech—from the early Internet to AI—to create real business value.

Over that time, I’ve developed an advanced protocol and toolset for innovating how to leverage AI and other digital technologies effectively. Before any experimentation is run or implementations are managed, there must be a phase of strategic rethinking and reinvention.

Even at the fastest pace possible, it takes at least six months—and usually over a year—to create the cognitive space and emotional safety needed for leaders to question entrenched notions about how their business truly works and challenge those assumptions with customer-centric insights that could unlock breakthrough innovations.

Without that space, teams default to tactical problem-solving: automating inefficiencies rather than rethinking the system. They rarely ask:

“What if our business model is outdated? What if our operating model is blocking AI innovation? What if our people model no longer fits how humans and AI collaborate best?”

True innovation requires slowing down before speeding up, engaging in strategic deliberation, gathering customer insights, and developing insight-driven concepts for how AI can solve problems better than existing solutions. This means you should pause, take stock, think strategically... and then act fast.

Yet many leaders want exponential results without the hard work of deep thinking, systemic strategizing, and genuine creative and conceptual thought. This is just one of many Innovation Killers that organizations develop to sabotage transformation before it begins. See The Ultimate List Of Innovation Mistakes, Failures & Killers.

Unless leaders protect the time, attention, brain power, courage, ambition, energy, and mindsets needed for strategic innovation and transformation, even the most advanced AI systems will merely optimize yesterday’s logic.

The 8 Phases of AI Innovation & digital Transformation Done Right

Successful leverage of AI for transformative impact requires progressive engagement in the opportunities and challenges. This might include steps like these:

1. Strategic Intent: Explore with truth and transparency what the true drivers for AI adoption are. Who and where is the demand coming from (Perceived desire from shareholders? Board members? The IT function? Frontline staff? What tensions exist (budgets, timescales, functions, etc)? What would good outcomes would look like.

2. Technical Insight: Understand what ML/AI actually is, what it can do well and where it often fails, the different kinds of ML/AI, the various value propositions and costs/benefits of existing offerings, where the industry is heading, etc.

3. Problem Definition: Reframe technical problems as innovation and transformation challenges that AI could solve, unlocking value in the process.

4. Customer Insight: Understand how the different types of AI can positively impact users, whether external customers or internal employees (this often impacts the business you think you're in). Begin with human-scale and customer-centric pain points that inspire energy and empathy to solve for.

5. Breakthrough Ideation & Concept Development: Use design thinking to generate new business, operating, and people models.

6. Experimentation & Proof of Concept: Test hypotheses through disciplined, low-risk experimentation and rapid prototyping to prove concepts before pilots.

7. Integration & Model Redesign: Redesign and reengineer processes, systems, incentives, and team behaviors for adoption and integration, including addressing potential risks and threats.

8. Scaling & Cultural Embedding: Build the culture and governance to enable AI transformation at scale and make it self-sustaining. This means deeply understanding hearts and minds and the often overlooked, but always critical, people side of digital transformation.

These are not theoretical. They are crucial for every successful AI innovation initiative and enterprise transformation program I’ve designed and guided for organizations around the world.

The key is to grok that the Pilots and Implementation/Integration that most people rush to—which leads to such poor ROI from AI—happens months after the creative and strategic thinking stages.

When leaders apply these transformation technologies alongside AI, there can be a multiplier effect. For example, AI can simulate, stimulate, and synthesize possibilities faster than humans can, but only humans can decide what transformations matter and where they should lead the organization.

I explore these synergies and possibilities in my AI, innovation, digital transformation, and leadership keynotes—where executives and teams can learn to blend human creativity and machine intelligence to deliver truly exponential value.

Book Nick Jankel, AI Keynote Speaker & AI Transformation Facilitator

Nick Jankel is a Cambridge-educated thought leader, a top 10 global keynote speaker, and an AI transformation advisor trusted by organizations like Microsoft, AbbVie, LEGO, and Unilever.

His AI keynote speeches and AI leadership programs ignite courageous, curious, and customer-centric minds and inspire meaningful and measurable change. Nick helps leaders and teams move beyond buzzwords to build businesses that truly harness the power of human + artificial intelligence.

INQUIRE NOW

What does an AI keynote speaker do?

An AI keynote speaker helps leaders and organizations understand how to use artificial intelligence as a catalyst for transformation — not just efficiency — inspiring teams to innovate and adapt for the future of work.

Why do most companies fail to get ROI from AI?

Most organizations retrofit AI into old workflows instead of rethinking their business and operating models. Real ROI comes from redesigning systems around what AI truly enables.

What are AI leadership programs?

AI leadership programs equip executives and teams to integrate human insight and machine intelligence, developing the creative, emotional, and strategic capacities needed for successful AI transformation. Visit SwitchOnLeadership.com to explore AI innovation sessions and AI leadership transformation programs.

How can AI improve leadership and innovation?

AI can accelerate innovation by simulating and stimulating new ideas, supporting creative problem-solving, and helping leaders envision and design smarter, more adaptive organizations.

How can I book Nick Jankel as an AI keynote speaker?

Use our contact form or email at office [at] nickjankel.com