Virtual café
The idea in The Little Prince that what matters is not always visible at first glance makes me think about how we use data to understand people. What interests me in that invitation to look again is recognising that what we can see may be only part of what we need to understand.
A CRM can tell us when someone contacted a company, what they asked or when they stopped replying. That is useful information. But it does not always explain what they expected, what made them hesitate or what had happened before. The record preserves what the system could collect and what someone decided was worth recording.
That selection deserves attention. If we only ask about what we already know how to measure, other questions may have no place. An expectation never expressed, or a difficulty that does not fit a form’s categories, may not appear in the analysis. Its absence from the table is not enough to conclude that it does not exist.
The same behaviour can also have different explanations. Leaving a purchase unfinished might relate to the price, an interruption or an unanswered question. These are possibilities, not conclusions about a particular person. The data points to something worth investigating. The explanation needs more context.
As I explore these tools, I wonder how much of what we call knowing someone consists of arranging what we have managed to record. A classification can help us work, but it may also stop us looking for what does not fit. I want to keep that reservation before turning a category into a complete description of someone.
AI can help find patterns and connect information. Even so, it works with the material available and the question we ask. If an experience was never recorded, a well-written answer may still fail to account for it. A convincing explanation does not resolve that gap in the information.
That is why I want to combine data with research and judgement. Listening to people can raise questions the initial analysis did not consider. A conversation cannot explain every experience either. We need to compare what emerges and remain willing to revise our interpretation.
Before assuming I understand a behaviour, I would like to be able to say which part I have observed and which part I am supposing. And know whom I need to listen to, to find out whether that explanation resembles what they experienced.