Data insights start, naturally, with data. Get the data wrong, and your insights will be wrong. Who wants to take business actions on insights based on bad data?
The real problem is when you’re acting on data insights that you don’t know are based on bad data. You need to prevent that from happening because you don’t want to make incorrect assumptions. If you’ve done nothing to build useful data, why should you expect it to be good?
The data industry has been around a long time and has proven techniques for building that solid foundation upon which to base your insights. Using these techniques and following a structured methodology provide the discipline needed to make that foundation. But building a good foundation that produces valuable insights takes time. Don’t try to cut corners and don’t assume your data is useful.
Another problem most companies face is that the business changed over time, but the infrastructure did not change with it. In older applications, for example, developers often built inflexible designs that could only handle the business processes at that time. As the business grew, started offering new products, or partnered with other companies, people found that the infrastructure needed to grow. Often, rather than taking the time to fix the infrastructure, they made compromises, sometimes for speed of implementation. The compromises don’t typically age well. Eventually, you need to take the time to fix it. Going forward, learn from past mistakes, and don’t make those compromises. Stick to your disciplined approach and don’t implement something until it’s ready.
continue reading »