> > I wonder if Amazon would check the Slack+Chime logs to see how the activity drops before and after RTO mandate.

> They already have and they found that productivity was way down. They don't care. They made this decision from the gut, despite claiming to be a data driven company. When anyone called Andy on it he just said "sometimes we make decisions that we know are right" blah blah.

Respectfully, do you know this because you yourself have seen the data, because a person you know saw this data and told you about it, or are you speculating that this is probably the case?

Because in a meeting with the CEO (which was publicly leaked so this isn't internal information I got it from outside sources myself I wasn't in the meeting) someone called out that the data doesn't support this decision and that was his response.

https://www.businessinsider.com/amazon-andy-jassy-no-data-re...

I don't see any claims in that article about productivity going up or down after the mandate, however? Just that data was not used to inform the decision in the first place.

Would love to hear knowledge of internal data on this either way but it sounds like we're still just guessing.

They have a tool at Amazon called "connections" which is a super annoying mandatory survey you must interact with every day. It measures a lot of things like whether our teams have certain internal processes in place, job satisfaction, tools satisfaction, etc. The surveys provide a list of fixed answers you can choose or if you rank something "negatively", you can select an option to describe why you ranked something poorly. Some of us have indirect access to the larger data through senior-level middle management who will either show us the connections data in private Slack DMs (if you're lucky enough to have a friend up that high) or they will blatantly tell us what the data shows in meetings. For example the job satisfaction one has been down below 50% satisfaction since RTO and management has explicitly stated it's due to RTO because that's what people are writing in as the reason.

Okay understood (that’s application sounds fascinating to be honest) and the drop in job satisfaction doesn’t surprise me at all.

But what about productivity/output?

Obviously this is partially a function of job satisfaction, or a correlation. But wouldn’t it be interesting if job satisfaction went down but productivity stayed the same?

> They made this decision from the gut, despite claiming to be a data driven company

I also work in a big "data driven company" and the same thing happened. Pretty sad that they bother engineers to justify everything they do with metrics (even when it's obviously wrong), but admitted in a Q&A that they didn't have productivity data to back up their RTO choice.

> Also they over-hired in the pandemic and need people to leave because their attrition numbers were too low for 2022 and 2023.

That's the real reason. Making people quit is cheaper than layoffs.