The vesting is very little the first 2 years. It’s only 5% year one and 15% year two.
I left amazon earlier this year. The way it works is year 3 and 4 are huge comp years if the stock has done well, since you vest 20% every 6 months. You make over the top of the pay range for your level.
Once you’re an employee the grants vest over 2 years. They’ll use the over top of pay range excuse to avoid giving raises or more stock. That means if you stay until year 5, then you’ll get a huge drop in comp.
Since you’re over the range, you’ll get told you need to be regularly seen in the office past 7 to get promoted. I was told my team wasn’t suffering enough.
The accomplishments on promotion emails are often terrible, but the people are always in the office. I saw a guy get promoted, because he did his evening grad school videos at work. Everyone thought he was working late and dedicated. I can’t name a single thing the guy ever delivered. I had two managers with horrific employee satisfaction scores get promoted. One was so bad her best question was below median. She was investigated by HR for verbal abuse, and she got promoted the next week.
The irony is when I left they had to add 2 individual contributors to the team in addition to my backfill to keep the pace. I decided to leverage the year 3 comp into a higher offer elsewhere.
The retail side is really, really mediocre talent wise. I didn’t work with anyone that had actually worked for a tech company. Most people were from brick and mortar retailers, transportation companies, industrial /automotive, or oil and gas.
A classmate’s friend was overseeing a UCLA MBA intern. This guy’s main concern was this intern wasn’t staying at work after 7pm and I was thinking in my mind “His main concern should be the interns quality of work and work output “
I had a colleague whose employee under him used to be at work all the time. He once complained to me in private “When my project needs I can’t even give him extra work“. Pretty sure he wasn’t that productive either.
Also, going home and logging back in doesn’t count. It needs to be in the office.
I did get insanely good at SQL. I’m working on migrating over to data science. I’ve done the data engineering work to get all the data normalized and stored in SQL. Now I need to learn how to apply ML to the data and create a predictive model. I leading the evaluation of learning platforms, so once one of those is selected I can take classes to learn the next steps.
I did my own regression model to determine which activities have the best correlation to results 1 and 2 months in advance. It’s super accurate. I’m curious if ML can really be more accurate.
Do not waste your time with ML and predictive model, it is not going to work. I tried around 25 open source LSTM, CNN and RNN, none has predictive power than my simple logic.
There was one reddit post long time back , one person worked appx 135+ open source predictive scripts, but none was working.
Some exclusive commercial algorithm may be working, but those are proprietary for the organization and mostly company secret.
This will work nicely. Use the one whichever is working , refine and refine until you get satisfactory results.
I connect excel to SQL. Then I applied linear regression in excel to all the inputs with a 1 month and 2 month lag vs output. I found one was 0.6 for 2 month and another was 0.6 for 1 month. The rest were garbage and some even negative. Then I did pretty simple math with the two that had highest correlation based on their indicated conversion rate vs input quantity.
I’ll let you know on the learning platform. We’re testing Udemy, Coursera, and Pluralsight. The top areas people want are: AWS, SQL, data analysis/visualization, Agile/scrum, and Jira. That’s how we selected people for the pilot and the training that’s being evaluated.
We aren’t evaluating it. It’s shocking how cheap enterprise plans are. It’s cheaper for unlimited classes than you’d pay for a single full-price course.
and those who said don’t compete with AMZN 
for cheap.