Skip to main content
THEN

Why organisations investing in AI are not yet seeing business outcomes and where executives should begin

A 2025 MIT study found that 95% of corporate AI projects fail to deliver measurable business results, despite over $30 billion invested worldwide. The root cause is not poor technology, but strategy, organisation, and people. Effective AI Transformation does not start with buying tools or running ChatGPT training; it starts with defining clear business challenges, measuring outcomes from day one, and preparing people. This article explains why most organisations are not seeing results yet and where executives should begin.

Hello, I would like to start with a scenario that many executives are likely familiar with: the organisation approves an AI budget, runs ChatGPT training for employees, and purchases several tools. Everyone is excited at first. But when the board asks, "So, how exactly has AI improved our business this year, and can you measure it in numbers?", the answer is vague—pointing only to the number of people trained and the number of tools purchased.

To be candid, this is not a problem unique to your organisation. Global research from MIT's 2025 The GenAI Divide study found that despite a combined investment of $30 to $40 billion in Generative AI, 95% of pilot projects fail to produce a measurable impact on the P&L, with only 5% delivering significant value.

And what the research highlights is that the difference between the successful 5% and the silent 95% lies not in the quality of the technology or the size of the budget, but in strategy, organisation, and culture. Put simply, effective AI Transformation does not start with tools, but with defining clear business challenges, measuring outcomes from day one, and preparing people. Today, I will guide you through where executives should begin.

Why are organisations investing in AI yet not seeing business results?

Because most organisations are stuck in the "experimentation" phase and fail to cross over into "actual transformation". There are numerous pilots, but projects never reach practical deployment that delivers results. MIT research terms this gap the GenAI Divide—the gap between organisations experimenting with AI and the select few that genuinely transform their business.

The numbers clearly confirm this picture. Although over 80% of organisations have experimented with tools like ChatGPT or Copilot, only around 5% of specialised AI solutions reach production deployment at a scale that generates ongoing value. And this issue is not improving over time. Data from S&P Global revealed that in 2025, 42% of organisations abandoned most of their AI projects, more than double the 17% recorded the previous year.

In Thailand, we are seeing the same signs. AWS's Unlocking Thailand's AI Potential 2026 report warns that Thailand risks a widening divide between organisations that merely experiment with AI and the few that genuinely transform how they work. Those who bridge this gap first will gain a long-term advantage.

What does effective AI Transformation start with?

It starts with business challenges, not tools. Organisations that fail often start with statements like, "We must have AI for sales," and rush to buy tools. Yet when looking closer, the actual bottleneck lies in fragmented data or unclear processes. Starting with tools addresses the wrong problem.

Executives who achieve effective AI Transformation typically begin with three steps: First, define clearly what business problem AI should solve and what results are expected. Second, establish metrics before starting, because research indicates that without baseline figures prior to AI adoption, there is no way to prove what has changed. Third, prepare people and processes for change, because AI delivers results only when people actually change the way they work.

A key point executives often overlook is that this is not solely an IT project. MIT research found that projects partnering with experienced external experts achieve significantly higher success rates than internal teams working alone. Effective AI Transformation is an organisation-wide agenda, not merely a software installation.

3 traps that leave AI investments yielding no results

Based on research and what I have observed in Thai organisations, there are three common traps:

Trap 1: Bolting AI onto existing processes Instead of redesigning processes to suit AI, organisations often overlay AI onto unadapted workflows. The result is a slight gain in speed without any structural transformation.

Trap 2: Investing in the wrong places MIT research found that over half of AI budgets are spent on flashy sales and marketing initiatives, even though applying AI to back-office operations yields much higher returns. Investments are chosen because they look impressive in boardrooms, not because they deliver tangible impact.

Trap 3: Treating it as an IT project rather than a business agenda When delegated solely to technical teams without business ownership or people readiness, projects stall at the pilot stage. This aligns with Gartner's forecast that 30% of Generative AI projects will be abandoned after the proof-of-concept stage by the end of 2025.

Assess which trap your organisation might be caught in

Before moving forward, I would like to invite executives to take a moment to evaluate: Are your organisation's current AI projects driven by business challenges or by tools? Were metrics defined from the outset? And is there active business ownership, or is it treated simply as an IT initiative?

If you wish to build an AI strategy that bridges the gap from experimentation to tangible results, Learning Hub Thailand offers dedicated executive courses. Explore our AI Strategy and Transformation for Executives course and full AI for Business course overview to design the right pathway for your organisation.

What are AI Agents and AI-Driven Enterprises, and why must executives understand them?

An AI Agent is an AI that autonomously executes multi-step workflows on behalf of humans to a certain degree—not merely answering one-off prompts like a chatbot, but receiving a goal and managing complex, multi-stage tasks. Meanwhile, an AI-Driven Enterprise is an organisation that designs its operational workflows with AI embedded from the outset, rather than treating it as an add-on tool.

Why must executives understand this now? Because this is the incoming next wave, yet it carries familiar risks. Gartner notes that over 40% of Agentic AI projects are likely to stall or be cancelled. The reason remains the same: launching without clear strategy and governance.

Executives do not need to write code themselves, but they must understand enough to define challenges, set boundaries, and establish governance frameworks. Courses such as AI Agent for Executive and Strategic AI Transformation are specifically designed for this purpose.

Fragmented AI Adoption vs Strategy-Driven AI: What is the difference?

Here is a clear comparison:

Fragmented AI Adoption

Strategy-Driven AI

Starts with "We must have AI" and searches for tools

Starts with business challenges and selects the right approach

Measured by number of people trained and tools purchased

Measured by business outcomes with upfront KPIs

Treated as an IT project

An executive and organisation-wide agenda

Budgets poured into high-visibility boardroom projects

Invested where tangible returns are delivered

Stuck at the pilot stage

Scaled into production and business-wide rollout

This difference marks the exact line of the GenAI Divide. Organisations on the wrong side waste budget and time on dead-end pilots, while those that bridge the gap are redesigning how they work and pulling ahead of competitors.

Workplace Story: From silent AI experiments to measurable business challenges

Let me share a common pattern I often encounter across multiple organisations, rather than any single individual.

Many executives share a similar story: last year, they invested heavily in AI, organised company-wide training, and purchased various tools. Yet by year-end, they could not demonstrate to the board how the business had transformed, because no one had set key metrics from the outset, and departments were merely experimenting in silos.

When they shifted their approach the following year, they did not start by buying more tools. Instead, they began with a single business problem with clear, measurable outcomes—such as reducing paperwork time in a department with high repetitive tasks. They set baseline metrics, assigned a business owner, and then selected the appropriate tools. The result was that this time, they could report back to the board with concrete numbers and use the initiative as a model to scale. The lesson is that effective AI Transformation does not start with the biggest budget, but with a clear, measurable problem.

Reflection Questions: Food for Thought

  • Did your AI initiative start with a business problem, or simply because "we need to have AI"?
  • Did you establish key performance indicators before launching the project?
  • If the board asks how AI benefits the business, can you answer with real figures?
  • Does the AI initiative have a business owner, or is it treated solely as an IT project?
  • Have you prepared your people and processes, or have you merely deployed new tools?

Next Action: Start Today

If you want AI to deliver real business impact, I suggest starting with one simple, actionable step: choose a single business problem that can be clearly measured and is important enough to interest the board. Establish a baseline figure today—such as time spent, cost incurred, or revenue generated—and then design how AI can assist.

By taking this step, you shift from investing in AI on vague hopes to running a project that proves its value. I believe that organisations achieving real impact with AI do not start with the most tools, but with leaders who define clear problems, measure outcomes effectively, and drive systemic and cultural change together. 💛

Are you ready to design an AI Transformation strategy that delivers real business results?

Learning Hub Thailand offers an AI for Business curriculum tailored specifically for executives, helping leadership teams formulate an AI strategy that moves beyond experimentation into tangible business impact, fully customised to your organisation's context and industry.

Note: Preparing your entire workforce for AI is the other half of success. Read more in our article How to Plan an Organisation-Wide AI Upskilling Strategy and explore leadership perspectives in the AI era in Leadership in the AI Era.

Frequently Asked Questions

Why do organisations invest in AI without seeing business results?

Because most organisations get trapped in the experimentation phase and fail to transition to impactful, live deployment. MIT research in 2025 found that 95% of AI initiatives do not produce measurable business results. The primary cause is not technology, but strategy, organisation, and culture—such as starting with tools rather than business problems, failing to set metrics, and treating AI as an IT project rather than a business agenda.

Where should an organisation begin its AI Transformation?

Start with business problems, not tools, by clearly defining what business issues AI should solve and what results are expected. Establish key performance indicators before starting, and prepare people and processes for change, because AI will only deliver results when people truly change the way they work, not just install software.

What is an AI Agent and how does it differ from a Chatbot?

An AI Agent is AI that takes a goal and autonomously manages multi-step tasks to a certain degree, unlike a Chatbot which responds one query at a time. Executives should understand AI Agents as they represent the next wave transforming workflows. However, there are risks: Gartner estimates that over 40% of Agentic AI projects are likely to stall or be cancelled if pursued without a clear strategy and governance.

Why is AI Transformation not just an IT task?

Because AI only creates value when it changes how people work and business processes, not merely by implementing technology. MIT research found that projects with business-side owners partnering with expert external partners have a far higher success rate than having internal IT teams build them alone. AI Transformation must therefore be on the agenda of executives and the entire organisation.

How can you measure the ROI of AI investments?

It can only be measured by establishing baseline metrics before adopting AI—such as time spent, costs, or volumes for chosen tasks—and comparing them after actual implementation. Research shows that many organisations cannot prove ROI because they never set baselines. Executives should therefore select problems with clearly measurable outcomes and define metrics before starting projects.

References: