SQL for Data Analysis

Chapter 9. Conclusion

Throughout the book, we’ve seen how SQL is a flexible and powerful language for a range of data analysis tasks. From data profiling to time series, text analysis, and anomaly detection, SQL can tackle a number of common requirements. Techniques and functions can also be combined in any given SQL statement to perform experiment analysis and build complex data sets. While SQL can’t accomplish all analysis goals, it fits well into the ecosystem of analysis tools.

In this final chapter, I’ll discuss a few additional types of analysis and point out how various SQL techniques covered in the book can be combined to accomplish them. Then I’ll wrap up with some resources that you can use to continue your journey of mastering data analysis or to dig deeper into specific topics.

Funnel Analysis

A funnel consists of a series of steps that must be completed to reach a defined goal. The goal might be registering for a service, completing a purchase, or obtaining a course completion certificate. Steps in a website purchase funnel, for example, might include clicking the “Add to Cart” button, filling out shipping information, entering a credit card, and finally clicking the “Place Order” button.

Funnel analysis combines elements of time series analysis, discussed in Chapter 3, and cohort analysis, discussed in Chapter 4. The data for funnel analysis comes from a time series of events, although in this case those events correspond to distinct real-world actions rather than ...