Auto Apply to hundreds of jobs in a few hours

Español

日本語

Data Analyst skills: what to learn first, and what can wait

You don't need to learn every tool on a data analyst posting before you apply. You need SQL that holds up under pressure, one dashboard tool you know well, and the habit of checking your numbers before anyone else does. Pay moves around, but skills take a while to build, so learn them in this order and let the rest wait.

2,633 open jobs

Step one

Gets you the interview

SQL joins, aggregation and window functionsKnow when a LEFT JOIN quietly doubles your revenue, how GROUP BY with HAVING works, and how ROW_NUMBER or LAG answers questions like each customer's first order. Practise on Postgres or BigQuery's public datasets, not just a tutorial sandbox.

Excel for fast checksPivot tables, XLOOKUP, SUMIFS and Power Query still handle a big share of the requests you'll get.

One dashboard tool, learned properlyPick Tableau or Power BI and go past drag and drop. In Tableau that means calculated fields and LOD expressions. In Power BI it means a clean star schema and basic DAX measures like CALCULATE. Recruiters filter on the tool name, so match it to the postings you want.

Cleaning messy data without losing rowsDuplicate customer IDs, dates stored as text, nulls that should be zero and nulls that shouldn't. Being able to say what you removed, and why, is what separates a real analyst from someone who just ran the query.

Step two

Gets you the offer

Defining a metric before you calculate itActive user, churned customer and conversion rate each have several reasonable definitions. Candidates who ask which one the business uses, and write it down, look like people who won't ship a number that marketing and finance argue about for a week.

Python with pandas for the jobs SQL handles badlyReshaping wide survey exports, merging files from a shared drive, or automating a report that someone rebuilds by hand every Monday. You don't need machine learning. A Jupyter notebook that loads, cleans, and charts data is enough for most teams.

Basic experiment readingKnow what an A/B test result actually says, what a confidence interval means in plain words, and why stopping a test early because it looks good is a mistake. Product and marketing teams lean on analysts for exactly this.

Writing the finding, not just the chartA short summary that opens with the answer, then the caveat, then the chart. Hiring managers often ask you to walk through a take-home, and the ones who get offers explain what should change on Monday because of it.

Step three

Gets you promoted

Modelling data with dbtBuilding tested, documented tables that other analysts query means you stop answering the same question several different ways. It's also how analysts move toward analytics engineering, which many teams treat as the senior path.

Knowing your warehouseUnderstanding how Snowflake or BigQuery charges for queries, why a partition filter matters, and where a table's data comes from lets you fix a slow dashboard yourself instead of opening a ticket with data engineering.

Owning a domain's numbersThe analyst who knows exactly how refunds hit the revenue table, or why the sales team's pipeline never matches the CRM export, becomes the person leaders ask first. That trust is what turns into a senior or lead title.

Pushing back on a requestSenior analysts turn "can you pull everything on churn" into one question with a decision attached. Learning to scope work that way, in a short intake doc or a quick call, is what lets you spend time on projects that matter.

Certificates worth your time

Certificate	Best for	Effort	Worth it?
Google Data Analytics Professional Certificate	Career changers who want a structured start	a couple of months of evenings	A decent map of the basics and a recognisable name on a resume. It won't carry you on its own, so pair it with a project on real data.
Microsoft Certified: Power BI Data Analyst Associate	Anyone targeting companies that run on Microsoft tools	a few weekends if you already use Power BI	The most useful one for working analysts. It tests DAX, data modelling and report design, which is the actual job at a Power BI shop.
Tableau Certified Data Analyst	Analysts whose target teams build in Tableau	a month or so of steady practice	Worth it if Tableau shows up in most of the postings you want. A strong Tableau Public profile can do the same job for free.

Exam content, prices and names get updated, so check the details on Google's, Microsoft's or Tableau's own site before you pay for anything.

Put it on your résumé like this

Weak

Created dashboards and reports using SQL and Tableau.

Strong

Rewrote 14 SQL reports as 3 dbt models feeding one Tableau dashboard, fixing a double-counted refund join that had overstated monthly revenue by $85K and cutting weekly report prep from 5 hours to 20 minutes.

Questions people ask

Should I learn Tableau or Power BI first?

Look at the postings in your area and count which one comes up more. Power BI is common at companies already on Microsoft tools, and Tableau shows up a lot in tech and marketing teams. Once you know one well, the other takes weeks, not months.

How much statistics does a data analyst really use?

Less than a stats degree covers, but more than averages. You'll use distributions, medians versus means, correlation and the logic behind a test result.

Is dbt worth learning before my first analyst job?

Not before. Learn it once you're writing SQL daily and notice the same logic copied across reports. That's the problem it solves, and it'll make sense right away at that point.

What skill do new analysts underrate most?

Checking row counts. Before and after every join, look at how many rows you have. It catches more wrong answers than any clever technique.

Got step one? Start applying. HeroApply matches you to roles that fit.

Start your trial