During World War II, the Allies wanted to reinforce the armor of their planes to reduce combat losses. They analyzed the aircraft returning from missions and identified the areas with the most bullet impacts. The initial conclusion was that these areas needed reinforcement.
However, a mathematician named Abraham Wald noticed a crucial mistake: the data came only from the planes that had managed to return, not from those that had been shot down. Wald interpreted the data from a different perspective: the planes that did not return were likely hit in critical areas such as the engine or the cockpit, which is why they never made it back. The solution was not to reinforce the parts with the most visible damage but rather those that had no damage on the planes that returned—because those were the areas where an impact meant total destruction.
This analysis is a classic example of survivorship bias, a common mistake in decision-making that focuses only on cases that have "survived" a process, ignoring those that have not.
This type of error does not only occur in military conflicts. Misinterpreting data often leads to poor decision-making. Data alone does not provide answers; the key lies in asking the right questions. This was the focus of the last Runroom LAB: Data + Product = Effective Strategy. In the LAB, Manuel Maffé,, Product Research Manager at HP, and Pablo Andrés Margara, Senior UX Researcher at Glovo, discussed how to interpret data strategically to avoid misleading ourselves and make better decisions.
The Trap of Collecting Data Without Purpose
Many companies are obsessed with collecting data without a clear purpose, hoping it will become useful at some point. Others chase metrics as their ultimate goal, failing to understand that metrics are merely tools to comprehend the context. When a metric becomes an objective, it loses its purpose. Without a research strategy, data is just noise.
In theory, making data-driven decisions sounds like the most rational and objective approach. But in practice, being 100% rational is impossible. There will always be variables we cannot measure, unknown information, and biases influencing our interpretation of the data. Intuition is not an obstacle to data-driven decision-making—it is what allows us to act when the data does not provide a definitive answer.
For years, intuition and data have been framed as opposing forces, but in reality, well-used intuition is accumulated experience. It enables us to recognize patterns. The question is not whether to choose between data or intuition but to understand that informed intuition is what truly drives innovation. To achieve this, we must wear different hats:
- 🔍 Archaeologists: We uncover hidden patterns in data.
- 📖 Historians: We understand how the past shapes the present.
- 🔭 Astronomers: We anticipate what is coming next.
The key is to combine these three perspectives—not just collecting data but transforming it into actionable knowledge.
Take the case of Steve Jobs and the launch of the iPhone. In 2007, the most popular phones were BlackBerrys with physical keyboards, and most market data suggested that users valued physical keyboards above all else. If Apple had relied solely on existing data, they would never have bet on a buttonless touchscreen. However, Jobs and his team intuited that touch interaction could revolutionize how people used mobile devices. Their intuition was not baseless—it was backed by observing technological trends and changes in user behavior.
The Key Insight
Data can tell us what is happening, but it cannot always tell us what to do. This is where informed intuition comes in—it allows us to take strategic risks based on a vision for the future.
From Data Collectors to Strategic Partners
Research and data analytics are evolving. Initially, research teams were mere data collectors. Then, they became knowledge providers. The next step is for researchers to become strategic partners in decision-making.
To achieve this, we must stop obsessing over superficial metrics and start building meaningful data:
- Select: Not everything that can be measured matters. We must identify what information is genuinely useful for business objectives.
- Validate: Data can be biased. It is crucial to question it and ensure it accurately reflects reality.
- Connect: Linking different information sources (market, brand, product) to gain a holistic view.
This approach ensures that data is genuinely useful and not just an accumulation of information sitting in dashboards no one checks. Data is not the answer; it is a tool for making better decisions. Collecting, storing, or analyzing it is not enough—the real challenge is giving it meaning and turning it into strategic action.
The story of the British bombers taught us that analyzing only what is visible can lead to poor decisions. Intuition showed us that sometimes, data alone is not enough to take the leap. The evolution of research is not just about improving analysis techniques but about changing our mindset—shifting from being data collectors to becoming architects of knowledge.
And in this transformation, the best researchers are not the ones with the most data but the ones who ask the best questions.