Another option is to use a cloudfront hosted pixel, and have a cloudwatch job to schedule processing of that data every X minutes. This gives you ultra fast edg

Another option is to use a cloudfront hosted pixel, and have a cloudwatch job to schedule processing of that data every X minutes. This gives you ultra fast edge response time on your tracker, instead of slow lambda + api gateway, but you have to back-process the data. This should also cost less if you have a lot of traffic. Instead of Google sheets, throw it to bigquery. Or reprocess hourly/daily into parquet then scan with athena.