Early adopters: Minimal Modeling for self-service analytics
Epoch8 are the early adopters of Minimal Modeling, and they are quite committed to it. They use Metabase and dbt extensively. Olga Tatarinova wrote a blog post on how they use Minimal Modeling for self-service analytics for their customers: “How minimal modeling can improve self-service analytics”.
For me, their implementation is about finding the common language between business users and the technical implementation:
“If I want to plot Daily Active USERS, why should I go to the EVENTS table?”
Concerns of data analysts and their customers’ needs is an interesting exercise in clarifying Minimal Modeling approach. Its original domain was more skewed towards the needs of software development and database maintenance, but now we see that simple and straightforward decomposition of datasets immensely helps with making sense of data.
Olga currently works on the guest post for this newsletter. Russian speakers (and Google Translate users?) can read their other case study: “End-to-end self-service marketing analytics for YouTravel”: https://agima.ai/youtravel.
Key takeaways from that:
“The dataset is fully documented, and updating the documentation is integral part of the process.
Thanks to the attribute independence, adding new data slices is easy.
As new people join the project, they can quickly get up to speed and begin contributing.”
Epoch8 could be reached at epoch8.co.