How Should Companies Get Started in Data Science?

 

At the end of the day, businesses recognize that measurement is crucial for improvement. The inability to measure aspects like costs, profits, and revenues hinders the ability to enhance or reduce them. Therefore, the initial step for a company is to start recording and capturing data, focusing on costs—differentiating them by labor and material costs, individual product costs, and total costs. Simultaneously, understanding revenue sources becomes vital, analyzing whether a significant portion comes from a small customer segment or vice versa.

The key advice is to initiate data capture. If not already capturing data, start doing so. If data is being captured, ensure proper archiving without overwriting old data, as historical data remains relevant. Consistency and documentation are essential for future understanding. Implement best practices for data archiving from the start or, if not done, prioritize it now.

Once data is in place, the next step is to apply algorithms and analytics. The importance of measuring cannot be overstated, as data science within a company is only as valuable as the quality of collected data. The principle of "garbage in, garbage out" underscores the significance of accurate and meaningful data.

For companies lacking existing data, the emphasis shifts to starting the collection process. Cleaning and analyzing existing data is equally crucial, ensuring its accuracy and relevance. The speakers stress the importance of having a team of engaged employees interested in data science. A team, rather than an individual data scientist, is highlighted as essential, each member contributing strengths in different areas of data science. Overall, the message emphasizes the fundamental role of data measurement and the strategic importance of building and maintaining a robust data infrastructure within a business.