i asked gpt-6 astra how to become a data analyst (it was wrong)

i asked gpt-6 astra for career advice (yikes)

September 24, 2026

Brought to you by: MyDataFolio

I created MyDataFolio after teaching how to build portfolios for 5 years and every other solution sucking.

Now there's an option that doesn't suck ;)

Everyone is freaking out about the new ChatGPT model Astra - apparently it's really, freakishly good. So I asked it how to become a data analyst.

The advice was actually pretty solid for the most part. But it said one thing that I think would pretty much ruin anyone's shot at landing a job.

So let me walk through the roadmap of becoming a data analyst - what it go right, and the really scary part.

What Astra got right

It did a good job on describing what data skills to learn:

Excel. Master pivot tables, VLOOKUP, and data visualization.

SQL. Learn to query databases, joins, aggregations.

Tableau or Power BI. Learn a BI tool.

I love this. This is exactly what I talk about with the Data Learning Ladder and what the data backs up on our skills report at FindADataJob.com.

These are the skills that are high-in-demand and pretty easy to learn.

It also said to build projects. But it didn't really give good examples how or emphasize publishing them enough. Doing projects is the best way to learn. But it's also the best way to convince hiring managers and recruiters how awesome you are. You need to publish them online.

Where it went really wrong

But lastly, and most importantly, it said you'll know when you're ready to apply for jobs when you feel confidence doing a full data project on your own.

UGH! :(

That's advice that keeps you from every landing a data job.

You'll never feel 100% ready to do a project on your own. You'll never feel ready. If you wait for that moment, you'll be stuck where you're at for years.

Instead, apply before you feel ready. Apply early and often.

Learn the skills, build the projects, and apply. But don't only apply - network. Because when you network you're more likely to land jobs.

That's how you actually land the job.

So use AI. But always double-check it. And stick to better sources to actually understand how to do complex things.

You're already doing that because you read this newsletter ;)

BTW, I made you a little app that creates you a data analyst roadmap based on your unique circumstances. It uses your real data, the SPN Method, and proven data to make you a real roadmap.

Try it free: DataAnalystRoadmap.com

LATEST EPISODE

i asked gpt-6 astra how to become a data analyst (it was wrong)

DATA JOBS OF THE WEEK

Junior Business Analyst Healthcare — Pyramid Consulting
Data Analyst Intern — CAI
Technical Business Analyst - Debt Solutions — CoStar Group
HP&S - Engineering Business Analyst *HYBRID* — RTX
Senior Associate - Senior Federal Financial Analyst — Blake Willson Group

👷How to build a data portfolio 100x faster than me:

I built my first data portfolio using GitHub Pages.

Here's everything I had to learn to build it:

GitHub: (repository configs, version control, branches, commits, deployment, etc)
Markdown language: (heading hashes, relative links, code blocks, etc)
Jekyll: (YAML files, static-sites, Liquid syntax, themes, etc)

I had to understand all 3 of those before even starting.

Took me hours - no, honestly, it took me weeks.

That's just to build the personal portfolio website - not even any of the projects.

Luckily for you - you don't have to do the same. I made a cheat code.