>>2105the infra costs are def the biggest hurdle when scaling to high traffic volumes. i've seen some setups where the cloud functions usage for sGTM basically doubled the monthly budget once we added more custom event streams. it's worth setting up
strict resource limits on your app engine instance early on so you don't get a massive surprise bill at the end of the month. managing the container configuration is also way more complex when you gotta handle things like user agent parsing manually.
>the real goal is long-term signal resiliencehow are you handling the data enrichment step on the server side without adding too much latency to the request? ❓