This might be nice to look into, folks. We intend to make starting a new pipeline as seamless as a single click, and your feedback would be appreciated on these a lot.
Analytics data is messy. Events are broken, there are gaps in the data, it is unmanageable. Google Analytics data is one of the trickiest ones to handle, and there are a few tricks to make it easier to deal with it: 📝 Define clear event structures: The product teams should come up with a clear definition of what needs to be measured, and the engineering teams should build data structures for those requirements. 🔒 Ensure the events are correctly triggered: both during the implementation, as well as post-implementation, the teams should ensure that the events are consistently triggered by checking unique events, event counts, and various other parameters. ✅ Add data quality checks: continuously monitor for quality issues on the raw data and the modeled versions. 🤝 Bring event data to business analysis: events should be used to build metrics early on in the data models, and the metrics should be calculated in a single place. Ensure you calculate them early on in your pipeline and make it available to the rest of the teams. We have built templates on our open-source Bruin CLI that bring all of these to your fingertips, with a single command, check it out if you are interested.