Hi, fellows! What is the best way to calculate computational (cloud server?) costs for a custom setup of FB Ads, Google Analytics, Google Ads -> BigQuery. I'd like to make a comparison between Stitch, Fivetran, Supermetrics. Their pricing is mostly based on queries and rows. What would the ETL computational costs be (excl. BigQuery costs) for 100 million rows or more? Could you give me a hint on where to look for what variables, so I can make the calculation?

@zlatan Hi Zlatan! The resource usage characteristics/requirements of a given pipeline will largely depend on the specific Singer tap and target used, and the overhead of Meltano's glue around and in between them is (should be) relatively limited. Taps and targets are inconsistently optimized for scale though, and not all follow best practices regarding parallelization etc, so you might want to ask around in the Singer Slack workspace if anyone's done this kind of analysis for the specific taps and target you're interested in using. I'm not aware of any aggregate data on popular tap performance, unfortunately, let alone in combination with Meltano.

Okay, got it. Thank you for the extensive answer 🙂