Found Data

You rely on numbers to make decisions every day, but most of those numbers were never intended for you. They were captured by sales teams recording purchases, or by servers tracking user clicks, or by sensors measuring temperature for equipment safety. When you use these numbers for your own goals, you are using 'found data'. Learning to distinguish between data designed for analysis and data found lying around is what allows you to judge whether your conclusions are real or just artifacts of how the records were kept. Without this distinction, every later decision about missing values, unusual entries, or system errors has no foundation.

Consider the difference between a diary you write to understand your own mind and an email you send to a friend. The diary is designed specifically to record your inner thoughts. The email was written to communicate a message, but if you read it years later, you might still learn something about what you were worrying about at the time. Found data is like that email: it was never created to be analyzed. It was created to run a business, to complete a sale, or to post on social media. Someone later says, 'These emails tell me about anxiety patterns,' or 'These sales records tell me about customer habits.' The data was found, not designed for that question. That origin leaves a mark on everything you do with it. The person who wrote the email had no idea you would read it for anxiety signals, so they did not write their true feelings clearly; they wrote what they wanted the recipient to see. Similarly, the salesperson who recorded a discount did not know you would later use that to predict demand; they recorded what the software required them to enter. The data carries the fingerprints of its original purpose, and you have to learn to see those fingerprints before you trust anything it tells you.

Imagine a city's surveillance cameras. They were installed to catch traffic violations, to monitor crowds, and to deter crime. Nobody installed them to study how people move through the city at different times of day. But a city planner comes along and says, 'I can use these camera feeds to understand pedestrian flow and design better sidewalks.' This is found data. The cameras now serve a second purpose they were never designed for. The crucial thing is that the way the cameras were placed biases what they see. They are dense in commercial areas and sparse in residential zones. A planner who forgets this will conclude that no one walks in residential areas, but that is an artifact of camera placement, not a fact about walking. The same happens in data. A social media platform records every click on the 'like' button but does not record the time someone hovers over a post without clicking. If you study engagement using only likes, you miss all the silent readers. The record was built for the platform's needs, not yours.

Found data is information that was generated for an original operational purpose—such as running a business or recording a transaction—and is later repurposed for analysis, carrying the biases of its original collection process into every conclusion drawn from it.

This concept sits beside Data Cleaning, which deals with the mess in real datasets like missing values and duplicates that often results from data being found rather than designed. It also connects to Data Validation Constraints, which are the rules you encode to handle expectations that found data so often violates.