16.18. Discussion: BigQuery and BigQuery ML Use Cases
Visualizing Recurring Revenue and Lifetime Value
Clients in the product or info product space often generate revenue through recurring membership fees, which means their focus is on metrics like lifetime value, churn rates, and stick rates. While it might not be necessary to use machine learning for these calculations, BigQuery can still help create valuable visualizations.
For instance, you can calculate a customer's lifetime value by simply summing up their purchases over time using SQL. If you want to calculate the number of transactions they made, just count the rows in your data.
Machine Learning Dimensions
If you're interested in incorporating machine learning into this process, explore different models available within BigQuery ML. Some models are great for calculating probabilities of an event occurring (e.g., a customer churning), while others excel at time series forecasting (e.g., predicting sales volumes).
For example, a linear regression model can help forecast future sales based on historical data while automatically accounting for seasonalities and holidays. This makes it easy to start with BigQuery ML - just provide your historical sales data and let the model do the rest.
Effective Visualization Techniques
While there may not be any specific examples of interesting visualizations that come to mind immediately, it's important to focus on the sequence of creating value:
- Ask questions and define use cases
- Model data to obtain necessary numbers
- Choose appropriate visualizations based on who will consume the report
The right visualization depends on who will use the report - some people might prefer tables while others need time series charts or other types of graphs.
By following this sequence and considering your audience's needs, you'll create more effective reports that drive insights and decision-making within your organization using Looker Studio along with BigQuery and BigQuery ML.