Yes, that is directly from the LinkedIn dashboard for Uber [1] which has company employment data. I am assuming that they are dependent on people listing their employment as Uber for it to be correct. I believe you have to be a Premium user to get the data.

Where did you get the 2x to 3x data point? Also, does your data source show anything different on the other issues around slowing of hiring and additional churn this year at Uber?

It is possible that the top line # might be problematic because it requires self reporting, but there is still signal in the # of people that are self reporting joining and leaving.

[1] https://www.linkedin.com/company-beta/1815218/

Probably a lot of Uber drivers, who are not employees, choose to list themselves as such.

> Yes, that is directly from the LinkedIn dashboard for Uber [1] which has company employment data. I am assuming that they are dependent on people listing their employment as Uber for it to be correct. I believe you have to be a Premium user to get the data.

You're sourcing Linkedin for employment data and using that to say there's a churn problem at Uber?

LinkedIn shows new hires and total employees. Churn can easily be calculated by [June Employees] + [June New Hires] - [July Employees]

Good plan. It's not like that number includes vendors, contractors, drivers or anything that could skew the numbers.

Sarcasm doesn’t really help to move a conversation forward.

At some point, you will learn to try to extract signal from imperfect data. It’s a key skill. If you wait for perfect data, you will miss the trends until they are in the rear view mirror.

Unless you believe that there has been a sudden mix shift in vendors vs. employees, the general analysis of churn and hiring still hold. You are only debating the multiplier on numerator and denominator.

Those are less important.

Why not filter the analysis only to profiles with keywords like"engineer" or "operations". Otherwise the analysis is unlikely to be directionally correct.