Open source feature flagging and experimentation platform for stress free deployments, measured impact, and smarter decisions.

GrowthBook Users

image.png

I'm thinking about query cost/performance optimization with the pre-compute dimensions setting. What are the tradeoffs I'm making by selecting pre-compute??

I'd anticipate fewer queries overall when precomputing dimensions. But is each query scanning more data? Or is it just more query compute? (something BigQuery doesn't charge for but may slow queries)

Each query will not scan more data as it only uses dimensions set up in your Exposure Query, but it will add a Group By to your query. This can increase query runtime and (if you don't just pay for rows scanned) costs, but it should be minimal as there are no additional joins or data, just more columns in a group by.

If you analyze by these dimensions even just a couple times an experiment, this option will almost certainly save you query costs.

Thanks. That's what I suspected. For us it's really just about speed, then. Since moving to using dashboards as my homebase for results views and including a bunch of dimensional cuts in there, I think it'll be a no brainer to switch it on.