A short guide to steps and must-know terminology Subscribe to my newsletter to learn more about product testing and data analysis. ✍️ Steps for conducting a product experiment: Formulate Hypothesis — why do you have to run the experiment, what is the ROI? Set Control and Variance: know your conversion rate (baseline) set the rate you expect — this is your Minimum Detectable Effect (MDE) …

How to break down user activity by day of the week in SQL Weekday usage analysis (also known as DOW analysis — day of the week) is my favorite quick data exploration. I often run it as the first query working on a new dataset — breaking user actions down on days of the week. …

How and when to apply revenue forecasting for data analysis — Financial forecasting is not only applicable for finance (as one might think). You also have to apply it for various product or marketing-related projects. It is the most common type of analysis and often functions as the foundation of getting growth rates for revenue data. You would use it to…

User engagement and frequency of usage analysis using SQL — User engagement is one of the best indicators that people love the product they use. You can improve user engagement if you can define and scale it. User engagement can mean any interaction with your product or it could also be a list of specific actions (or that one action). …

Common product test mistakes and a practical guide for successful A/B tests — Many companies today are very aggressive in running product tests. Some make testing part of their culture, and some simply use it as a security tool to validate new features. …

Advanced SQL challenge on how to generate a time series in Postgres or RedShift. — Let’s say you have a task to make a histogram or a timeline of DAU (Daily Active Users) or users who purchased or subscribed daily, or any graph based on date values: You’d need to prepare the data in the following format that would have a date and a count…

How to read and run correlation plots in Python Pandas — Do you ever receive questions like: Does correlation imply causation? How do you prove if features X and Y are correlated? After this helpful guide, you will know the best way to answer those types of questions. …

Learn how to make a word cloud for text analysis — Word or text clouds are very common tasks for analysts who work with textural, qualitative, or semantical data analysis. They are also common take-home assignments for candidates to test their knowledge of handling, processing, and visualizing text data. …

Case study: how increasing user activity affects company revenue — Product Analysts are often asked to quantify the relationship between user actions and company growth. This case is a common challenge for analysts and product managers to figure out how improving user retention and optimizing product flow or features leads to increased revenue. …