I was ready to look past the title as just the cost of doing "business" in 2019, but the article has a lot of problems. Stuff like measuring latency from 1 com

I was ready to look past the title as just the cost of doing "business" in 2019, but the article has a lot of problems. Stuff like measuring latency from 1 computer in Iceland -- gaining empirical results from your actually running user-base (logs & analytics) would be a much better test, as it is you can probably effectively ignore that part of the article because that's not an effective test. 8x multiplier on ~1K a month ignores the actual value proposition of serverless. Which is that you're not spending staff hours designing, building, and maintaining scalable infrastructure. If you don't have scalability issues (which at 1K a month, you don't), then it's maybe not the right technology for your use case. If you do, then maybe it's the right technology -- yeah you're paying more but you gotta do the math on what you'd need to build and then discuss tradeoffs with the business. Classic technology cost analysis, this is not new.