We often hear the phrase “data is the new oil.” It sounds powerful, it sticks in people’s minds, and it seems to make sense at first glance. After all, just as oil fueled the industrial age, data appears to fuel the age of artificial intelligence. But is this comparison really true or is it just a catchy slogan?

In this blog article in collaboration with Zsofia Raffa, we unpack the myths and realities of data in the age of AI. Together, we explore why this topic matters for leaders, the most common mistakes businesses make, and how organisations can take practical steps to turn data into a real strategic advantage.

 

Why Do People Call Data “the New Oil”?

 

The analogy between data and oil exists for a reason. Without oil, machines could not run, and without data, AI systems cannot function. Just as crude oil needs to be refined before it becomes usable, raw data must be cleaned, structured, and processed before it holds real value.

However, there is one crucial difference. Oil is finite. Once it is consumed, it is gone. Data, on the other hand, can be reused, enriched, shared, and combined with other sources. In fact, data often increases in value the more it is connected and contextualized.

Takeaway: While the “oil” metaphor helps illustrate the importance of data, it is ultimately incomplete. Data behaves more like renewable energy than oil—it grows, evolves, and generates new value over time.

 

 

The Difference Between Good Data and Bad Data

 

When it comes to artificial intelligence, quality matters far more than quantity. Good data is clean, accurate, consistent, relevant, and representative. It provides the foundation for AI systems to deliver reliable insights and predictions.

Bad data, in contrast, is incomplete, outdated, biased, or mislabeled. Feeding bad data into an AI system is like asking a chef to prepare a gourmet meal using spoiled ingredients. No matter how advanced the algorithm, the result will be disappointing at best and harmful at worst.

Takeaway: The success of AI does not depend on having more data. It depends on having the right data.

 

 

The Most Common Mistakes Companies Make

 

Despite investing heavily in data and AI, many companies still struggle to generate meaningful outcomes. The problem is often not the technology, but the way data is managed.

The most common mistakes include:

• Data silos: Departments guard their own datasets without sharing them across the organization. This prevents leaders from seeing the full picture. Imagine sales and customer support working with separate data—they might both miss crucial signals about customer churn.
• No clear strategy: Organizations collect data simply because they can, not because it supports a defined decision or business goal. This leads to rising storage costs but very little business value.
• Poor data hygiene: Datasets are riddled with duplicates, missing values, inconsistent formats, or outdated records. Leaders then make decisions based on incomplete or misleading information, which erodes trust.
• Quantity over quality: There is a persistent belief that “more data is always better.” In reality, irrelevant or biased data not only wastes resources, it actively harms AI models by diluting accuracy.

Takeaway: The biggest mistake is treating data as an afterthought. Data must be treated as a strategic asset—curated, governed, and aligned with clear business objectives.

 

 

Why Starting Small Can Create Big Impact

 

Many leaders are tempted to launch grand AI transformation projects. They aim for moonshot initiatives that promise to change everything at once. In reality, this approach is too complex, too risky, and too slow to deliver measurable results.

The smarter strategy is to start small and scale with purpose. This means focusing on one high-value use case that is both visible to the business and achievable in a short timeframe.

Examples include:

• Automating a time-consuming weekly report to free up employee capacity.
• Predicting supply chain disruptions before they occur.
• Using AI to categorise customer feedback and improve response times.

These kinds of “quick wins” prove that AI delivers real business value. They build trust with stakeholders, secure executive support, and create momentum for scaling.

Takeaway: Think of AI adoption as a journey, not a leap. By starting small, organisations build credibility and confidence that fuel long-term transformation.

 

 

The Critical Role of Data Ethics

 

Data ethics is not just about compliance—it is about trust. When organisations fail to act ethically, they risk damaging their reputation, alienating customers, and facing costly regulatory consequences.

There are three areas where ethical considerations matter most:

1. Bias and fairness: If training data contains hidden biases, AI systems will replicate and amplify them. This can lead to discriminatory hiring, unfair lending practices, or biased healthcare recommendations.
2. Privacy and consent: Customers expect transparency and responsible handling of their personal information. Using data without consent not only risks regulatory penalties under laws like GDPR, but it also destroys trust.
3. Accountability and transparency: People deserve to know how decisions that affect them are made. Black-box AI systems erode confidence, while explainable models help users understand and trust outcomes.

Takeaway: Ethics is not optional. It is the foundation for compliance, customer loyalty, and long-term success. Companies that embed ethics into their data strategy gain a competitive advantage by building stronger, more trustworthy relationships.

 

 

Data Governance as the Foundation

 

Without strong governance, data quickly becomes unreliable and unmanageable. Many organisations suffer from duplicated datasets, unclear ownership, and inconsistent standards. This chaos weakens confidence in both the data itself and the AI solutions built on top of it.

Effective governance provides clarity and accountability. It rests on four pillars:

• People: Clear ownership and stewardship responsibilities ensure accountability for data quality.
• Processes: Defined standards for how data is created, validated, cleaned, updated, and shared.
• Policies: Rules that guarantee privacy, security, and compliance with regulations such as the EU AI Act and GDPR.
• Tools: Modern platforms that make governance scalable by cataloguing, tracking, and securing data.

Strong governance does not slow innovation—it accelerates it. When teams can trust their data, they innovate faster. Leaders can make decisions with confidence, and AI initiatives can scale sustainably.

Takeaway: Governance transforms data from raw information into a trusted business asset. It is the foundation that allows AI initiatives to succeed at scale.

 

 

One Piece of Advice for Leaders

 

If there is one message to remember, it is this: do not wait for perfect data. Perfection will never come. Start with what you have, learn quickly, and improve along the way.

Most importantly, treat data as a product rather than a byproduct. Design it with the same intention you would design a product for customers, useful, reliable, and valuable.

 

I would love to hear your perspective: Do you think “data is the new oil” is still a helpful analogy, or do we need a new way of thinking about data in the age of AI?