Rant...

Question

Entry level DA for a small business. I'm doing adhoc reports from management and it's driving me up the fucking wall. Sales, transactions, etc, by age group, by store, etc. I even did one by gender, which we don't collect data on and took me so much fucking time to populate. Do you know how many names are spelled all so differently and I'm the one who has to guess their fucking gender and look them up one by one on the internet or simply guess. I have like 12,000 unique names on a fucking list right now. Now they want the two merged together. It's like, bruh, what fucking business decisions are you going to make by know the percent breakdown of consumed product by age group by gender? It's all fucking busy work.

TLDR; looking for a new job.

Answer 1

> what fucking business decisions are you going to make by know …

If you aren’t already asking that question when you get asked to do a new task, start asking (politely). I ask this of my stakeholders all the time. “What question will this help you answer?” Sometimes it’s to get out of useless analysis, other times the answer will help me make sure I’m looking at the data correctly.

Additionally, who do you report to? Start enlisting them with help for prioritizing your work. Keep a list of all of your tasks/projects, and review them in a weekly 1:1 (or more frequently). If you have more work than you have time to do, ask your boss what can come off of your plate or get assigned to someone else or just isn’t worth doing. And your boss can fight the battle if someone is mad you aren’t working on something.

Also consider some kind of intake process for requests. My team uses JIRA. Sometimes just making them fill out a form helps weed out the stuff that isn’t a priory. But it also helps you quantify how much work you’re actually doing or being asked to do.

All of the above is also how teams justify hiring more people. If they are asked to do more work than they have bandwidth for *and* the work is important.

Answer 2

did you assume their gender ...jk.. i know the feeling, i am working with different departments on naming conventions and they dont understand why that could be important and they come with exceptions all the time which have to be implemented manually....

Answer 3

There are packages/libraries to identify gender by name and if it's not possible, it gives you a probability.

Answer 4

For the name issue, I would personally get a bunch of lists of male and female names from those stupid name websites, combine them all together with male and females being in separate columns, remove the duplicates and populate a table with it in the database. Then, you could just use a  SQL case/when statement for boy/girl depending upon whether the first name field is in those respective lists. That would knock out like 80% of the gender labeling. Then, I’d just ignore the null values for now and gradually add the more creative names to my table to improve the accuracy.

This all goes without saying that this is still kind of silly because you can have a female with a first name Dale or something.

Answer 5

Is this real?

Get a list of baby names with their gender and join it.

Answer 6

Learn how to build an OLAP data cube and then set it up. It will take you some up front effort, but once it is set up, you can answer most of these questions in 2 Minutes.

Answer 7

There’s lists available with this information in it. R has RandomNames package. Im sure Python has an equivalent.

Answer 8

Python has a gender guesser library which you may find useful. But yeah this sorta thing sucks

Answer 9

It's not only asinine to assume gender by name, it's unethical and borderline illegal depending on what action gets done with the info. This is NOT a shot at you! Just a shot at whoever is asking you to do that lol!

Answer 10

If you don't collect data on something you can't make reports of it.

Trying to guess or work with some package/library is not an efficient use of time.

It is incredibly important to know consumed product by age and gender, that is like the most common metric to look at.

Seems like you need to take a step back and think things through, and ask questions before doing things and getting upset.