Hey everyone, I’m curious how others approach fact table management! Right now, we have separate fact tables set up for each of our brands. We originally went this route to minimize query costs — by avoiding scanning datasets not relevant to the experiment. That worked well from a cost perspective, but we’ve hit a bottleneck in maintaining metrics. Since our brands all follow the same schema (GA4 events), any new metric — say, a purchase event — needs to be added to each fact table separately. The same goes for updating the underlying queries. It’s a lot of manual work that could be avoided with a more centralized setup. So I’ve started exploring the idea of combining all datasets into one unified fact table. This would allow us to define a metric once and reuse it across brands, making maintenance much easier. The trade-off, of course, is increased query costs and potential performance implications. Curious to hear how have you handled this? Have you centralized your fact tables, or stuck with one per brand? And how have you weighed cost vs. maintainability?

We don't have the same problem but I guess partitioning and clustering in BigQuery could help keeping the cost down for a unified table

Thanks for that tip @adorable-bear-66287! Will definitely take that with me