The Paradox of AI Investment: BCG's Executive Radar 2025

Companies are pouring massive amounts of money into artificial intelligence, but most aren't seeing the returns they expect. In economics, the Jevons Paradox occurs when technological advancements make a resource more efficient to use; however, as the cost of using the resource drops this results in overall demand increasing, causing total resource consumption to rise. 

This counterintuitive effect demonstrates that efficiency alone doesn't always reduce overall expenses, and a recent study by the Boston Consulting Group reveals a sobering parallel with AI investments across global markets.

The Investment Gap is Lagging Behind

Only 25% of companies investing in AI are seeing meaningful returns on their AI investment. This statistic is particularly striking given that many of these organisations are spending upwards of $25+ million annually on artificial intelligence initiatives. The gap between investment and deliverable results suggests fundamental problems in how companies are approaching AI adoption.

Several factors are identified in the survey as contributing to these disappointing outcomes, ranging from executive leadership that lacks understanding of modern tech to individual and department-level mismanagement. Many organisations struggle with outdated workflows that weren't designed for AI integration and data science teams frequently over-engineer solutions, prioritising technical accuracy over business value. Perhaps most critically, independently siloed teams with misaligned goals and objectives create barriers to successful implementation.

These leading companies in the report maximise value creation through

🤝People and processes: invest in employee training and improve existing efficiencies

🏗️Focusing on outcomes: not just the possibilities of deploying a new technology

⛳Fewer AI programmes: commit to high-impact depth over breadth implementations

What Success Looks Like

Those 25% of companies that achieve a meaningful ROI follow a different playbook entirely. They focus relentlessly on business outcomes rather than getting caught up in technological complexity. These successful organisations select fewer use cases but delve deeper into each one, ensuring a substantial impact.

The study highlighted the 10-20-70 principle, where the majority of investment and effort is in People and Processes. Successful organisations redesign their workflows comprehensively, rather than simply layering AI onto existing processes. Instead, they reimagine how work gets done and apply AI as a fundamental consideration in their operations, upskilling their people and overhauling their process design to capitalise on AI opportunities.

The guiding principles of adoption should centre around:

🧬Prioritising core value: AI should be delivering on actual business activity

↪️Redesigning workflows: align business metrics and incentives end-to-end

🧷Break organisational silos: integrate business, tech, and data teams together

Pathways Towards the Future

The companies that succeed will be those that learn from the 25% already seeing returns, focusing on core processes and redesigning workflows rather than simply adding AI blindly and hoping for the best. The AI revolution is real, but success requires more than spending large amounts of money to have the most powerful tech; it demands fundamental changes in how organisations think, work, and measure value.

AI requires a deep reimagination of work and is not a silver bullet for productivity impact or cost savings. While AI offers huge potential for efficiency, using it to reduce headcount can create hidden risks, as rushing to replace people with automation often leads to drops in service quality, customer frustration and team morale. AI struggles with complex decisions, human nuance, and emotional intelligence which can result in a short-term cost saving that damages long-term trust, performance, and brand reputation.

Instead, the smartest organisations use AI to augment staff, not replace them. When deployed to eliminate repetitive tasks and free up skilled staff for higher-value work, AI becomes a catalyst for growth, not a hasty shortcut to downsizing. Only 7% of executives surveyed viewed AI automation as leading to a decrease in workforce, with over three-quarters looking to expand or upskill their organisations to implement AI skills.

At Lida, we help businesses strike the right balance: delivering measurable efficiency while keeping people at the heart of service delivery. Let’s talk about how to make AI work for you without the costly setbacks. Check out our AI Roadmap for more information or get in touch.