What to learn next
You’ve finished a 5-day intensive. That gets you to fluent-beginner — you can answer a real business question on a real dataset and ship a stakeholder report. The next 5 days, 5 weeks, or 5 months depend on where you’re headed.
If your next role is “data analyst” — go deeper, not wider
You have breadth from this course. Pick one tool and go deep before adding more.
| Tool | One book, one course, one habit |
|---|---|
| SQL | Book: SQL for Data Analysis (Cathy Tanimura). Course: Mode’s SQL Tutorial (free). Habit: write at least one SQL query every working day for a month. |
| pandas / Python | Book: Python for Data Analysis (Wes McKinney, the pandas author). Course: DataCamp’s Data Analyst with Python track. Habit: re-implement one of your team’s recurring SQL reports as a pandas notebook. |
| Power BI | Course: Microsoft’s PL-300 learning path (free). Habit: rebuild a dashboard your company already uses, then ask whoever made the original what they’d change. |
Statistical thinking — the next layer above the tools
This course is light on statistics on purpose. A real analyst eventually needs:
- Descriptive statistics done well — when to use mean vs median, what standard deviation actually tells you, how to spot a skewed distribution before it skews your conclusion.
- Correlation vs causation — why “X correlates with Y” almost never means “X causes Y,” and what extra evidence you’d need.
- Sampling and confidence — how big a sample do you need, what does a 95% confidence interval mean, when is a difference real vs noise.
- A/B testing fundamentals — if you’ll work at a company that ships experiments, this becomes daily vocabulary.
Two good starting points:
- Book: Think Stats (Allen Downey) — free at greenteapress.com.
- Course: DataCamp’s “Statistical Thinking in Python” or Khan Academy’s intro statistics course (both free).
Data visualisation — beyond Power BI mechanics
Day 4 taught you which Power BI visuals exist and how to wire them up. What it didn’t fully teach: which chart to pick for which question, what makes a chart honest, how to design for the reader. Two recommendations:
- Book: Storytelling with Data (Cole Nussbaumer Knaflic). 200 pages, every page useful, single best book on this topic.
- Reference card: The Financial Times’ Visual Vocabulary. A one-page map of “this question → this chart.”
AI-assisted analytics — going further than Day 5
Day 5 introduced Claude Code for analytics. If that resonated:
- Practice: use Claude Code for your next three real analyses (at work or on a public dataset). Verify everything. Notice when it saves you time and when it costs you time.
- Read: Anthropic’s Claude Code documentation — especially the “best practices” guide.
- Explore: other AI-coding tools (Cursor, GitHub Copilot) — they share the same fundamentals; prompts that work for one mostly transfer.
A portfolio of public work
Hiring managers read your GitHub. Three reasonable ways to build a public portfolio after this course:
- Re-do the capstone with a different dataset. Same business-question framing, different domain (healthcare, finance, sports). Pick something you actually care about — your output reads differently when you’re curious.
- Find a dataset on Kaggle and write a Day-5-style 1-page report. No Kaggle kernel, no leaderboard — just the markdown report and the underlying CSV / SQL / notebook in a public repo.
- Reproduce a published analysis. Take a finding from a news article (often based on public data) and try to recreate the numbers from scratch. You’ll learn more from one reproduction than from ten tutorials.
What we don’t recommend (yet)
| Skip for now | Why |
|---|---|
| Machine learning / scikit-learn / TensorFlow | Different role (data scientist). Get tool-fluent and stats-fluent first; ML is the layer above. |
| Big data tools (Spark, Hadoop) | Until your company hits “this CSV is 10GB,” pandas + SQL covers everything. |
| Dashboards-as-a-service platforms (Looker, Mode, Hex) | The mental model is the same as Power BI; learn one platform deeply before touching the others. |
A last note
The fastest way to keep learning after this course is to answer real questions for real people who care about the answers. Volunteer to pull a number for a friend’s small business. Take on a “this should be easy, can you just…” request at work. The difference between someone who took an analytics course and someone who does analytics is who replied yes to that next request.