Data-Informed, NOT Data-Driven — Ant Murphy

The post argues that being data-driven is a trap because data is never fully unbiased. Selection bias, recency bias, and confirmation bias can shape both the data itself and the way people interpret it, so data should support judgment rather than replace it.

It recommends using data to inform decisions, not to make them blindly. The author suggests working cross-functionally, testing ideas as if they are wrong, collecting multiple data points, reflecting on past decisions, and learning to recognize bias. He also says AI can help synthesize data, but its output still depends on the quality of the input and on human strategy.

Reading notes

#

• Data can be useful, but it is not objective or perfect.
• Selection bias, recency bias, and confirmation bias affect gathering, interpretation, and conclusions.
• Data-driven operating can turn into rationalization backed by spreadsheets.
• Data should improve decision quality, not replace qualitative judgment and user empathy.
• Cross-functional work helps avoid siloed perspectives.
• Discovery should test ideas as if they are wrong, not as if they are already right.
• Combining qualitative and quantitative data helps expose weak conclusions.
• Reviewing past decisions can reveal where data helped or misled.
• Awareness of biases helps identify when conclusions are being distorted.
• AI can synthesize data, but the result still depends on the quality of the data and the need for strategy.