Doesn't make any sense, the computation time difference between pulling data from one system and fifty systems isn't measured in the months range.

It makes a difference if it's just humans doing the job. Given the long window I would assume that this lacks automation

Yes, but "having a human in the system" isn't answering the original question: what does Amazon benefit from introducing excess cost into this equation? Surely there should be nothing clandestine in your own usage data, search queries etc.

> Yes, but "having a human in the system" isn't answering the original question

It does. There is a human manually running queries to copy your data into an Excel spreadsheet (or equivalent) then passing that onto the next part in the chain. This is done quite a few times and stalls the process.

> what does Amazon benefit from introducing excess cost into this equation?

It's cheaper to have the already existing employees do this. They haven't hire a new team to do this.

> Surely there should be nothing clandestine in your own usage data, search queries etc.

I suspect there's more than you own personal data mixed into these systems. Your personal data is mixed in with Amazon's company confidential data so they need to separate these out before sending you only what's required by law.

Thanks. I still find it surprising a company of Amazon's stature wouldn't have figured out how to fully automate something relatively trivial like this.