Part 1: Events

Event is the unit of data

Each event represents a tiny piece of information about the product, about a user action, or something happening in the background related to the user.

All analytics is built on top of these events. When we put events together, we assemble a complete picture of what's happening in our mobile app or website.

An AI-generated image based on the above lines.

As analysts, we are always complaining about events - the names don't make sense, the fields are all wrong, the triggers are poorly done, and there are too many (or too few) events. The list is endless.

How many times you have replied with 'the events are broken / unavailable' to a data or analysis request in the past month?

Yet, most of us haven't thought about them in enough detail. Analysts only work with events once they are in a table in their data warehouse.

As luck would have it, I have worked on event implementation for years and finally developed a set of best practices that allowed me to get rid of most of these complaints.

Good events save hours every week, reduce errors, and make querying intuitive. They also help with onboarding new hires.

It is hard, unrewarding work (only other analysts will praise you for good events). The good part is once it's done - you can reap the benefits for years!

This first module is all about events. We will cover -

• event components and how to get the most from them

• common event sources like client, server, 3rd party tools for attribution and payments

• how to create an event plan for a new feature

• best practices for naming events and fields and storing different types of values

• too little and too many events (event volume and iteration)

• moving events from a source to a data warehouse (data transfer)

• maintaining a spreadsheet with details about events (data dictionary)

We will end the module with a webinar where we will take a mobile app and a website and create an event plan for them.