ZOD: Two years as a Data Analyst... Here's What I've learned
Zen of Data Science #17 (Special Issue)
The Big Idea(s)
I’ve been a data analyst for a little over two years, and in this special issue I wanna share some of the top lessons I’ve learned.
(A lot of these apply to any field in data, or even programming.)
Tools are not most important… domain knowledge is.
It’s great if you know SQL, Tableau, Power BI, Python, R and on an on… but what can you do with it? I think many people, myself included, focus too much on the tools, and not enough on the business aspect of the job.
If you have advanced machine learning skills, but can’t apply them to the business you work for, why would they care? But if you create a simple report in Excel that saves the company 500k dollars a year… they’ll think you’re a hero.
Don’t just focus on learning tools. Study business, and your specific industry, and think how those tools can make an impact.
The learning never stops.
The more I learn, the less I know.
Over the last two years I’ve gotten very comfortable with SQL, and a few other tools, but I’ve realized how much I don’t know about others.
Nowadays it seems like companies expect a Data Analyst to know stats, machine learning, programming… it’s getting insane.
It takes a lot to stay relevant in tech. I put in a lot of hours of studying on top of my job. Partly because I enjoy it, partly because I don’t want to fade into obscurity.
Ask for help the right way
Early in my journey, when I was stuck I would ask my manager for help, but I wouldn’t come with a lot of information.
I would say,
“I can’t figure out where to find this data. Is it even in the source data base or should I be looking somewhere else?”
Which would lead to A LOT of back and forth.
Nowadays, if I was in the same position I would say something like:
“Hey manager, I can’t find this data mentioned in the Jira ticket [Link to ticket]. I’ve checked database a, b, and source data, and I even looked on Confluence [link to page], but haven’t been able to find it. Am I missing something here?”
A lot more thorough, and lets my manager know I’ve actually tried to find or solve something myself. It also prevents them from having to tell me “Well try checking database a, and b.” when I’ve already done that.
You get the point.
The Impact of AI is overblown
When AI first came out, I thought it was really cool. Then I slowly started to fear for my job as it got better and better.
There was even a point when I was considering transitioning out of tech, because I believed my job would become obsolete in a few years… (not kidding)
But as time goes on, and I’ve stress tested AI more and more for my own work, I’m convinced it’s nothing more than a tool.
I have asked AI to write code, or help me modify current code, and even with very, VERY, detailed prompting, it has given me hot garbage in return.
I asked it to write me a query a few weeks back, and it gave me a query that took about 210MB to run. I wrote the same query myself, and it cost 20MB to run… I then asked ChatGPT, Claude, and Gemini, to rewrite the AI query to make it more cost efficient… They all gave me basically the same thing, which still cost over 200MB to run.
Understand that this isn’t a huge deal for this specific query, but that is 10X more expensive of a query. If we had a query that cost $100 to run, then the AI generated version would’ve cost $1000 dollars to run…
People who say our jobs will be replaced by AI really don’t understand our jobs, or they’re doing very simple work.
If we’re replaced by AI, so will just about every other job out there.
Study what interests you.
Lastly, try to spend time learning what you’re genuinely interested in, even if you don’t see how it can help in your current job.
For a long time I stopped studying Python, because I don’t use it in my job.
I also stopped building personal projects because honestly? I find data projects boring as hell. I don’t want to do a whole case study on the cost of avocados, or video game sales.
But recently, I’ve started building projects just for fun, and learning what I feel like learning.
The last few months of doing this are the best I’ve felt in a long time. I study Python, and build a little app. I study using The Odin Project, and build a website using HTML/CSS. Or more recently I built a dashboard using Looker, to track how many jobs I’ve applied to, rejections I’ve gotten, interviews, offers, etc.
Building projects you actually care about is the best way to learn.
Of course sometimes you’ve gotta study or learn things you’re not crazy about, but whenever possible, try to tie them to what you’re genuinely interested in.
You’ll learn more, retain more, and genuinely have fun in the process.
It’s been a good first few years, and I can’t wait to see what the next few years bring.
Thanks for reading, and have a great week!
– Josh