Sefism Data Analytics Cheat Sheet

What this cheat sheet is

The Data analytics cheat sheet is the quick-revision companion to Sefism's Data analytics roadmap. It covers the ideas behind each stage, from the first business question to a dashboard and a clear recommendation: spreadsheets, SQL, statistics, your analysis tool, cleaning and modelling, charts and presenting, and your first analyst role.

Every concept explains what it is, why it matters and the classic mistake, and many come with a worked example or a snippet: spreadsheet formulas, SQL queries, DAX measures, Tableau calculations or pandas code, depending on your edition. Thirty-two concepts across the sheet include code.

Three editions

The edition follows the tool you chose on the roadmap. Each one has eight chapters: seven shared and one for your tool.

• Power BI edition (73 concepts): Power Query, relationships and filter direction, DAX measures, CALCULATE and filter context, date tables and time intelligence, report pages and slicers, and publishing safely.
• Tableau edition (73 concepts): data sources, relationships and joins, dimensions and measures, discrete and continuous, calculated fields, LOD expressions, table calculations, dashboards and actions, and publishing to Tableau Public.
• Python edition (74 concepts): Jupyter notebooks, the pandas DataFrame, selecting and filtering, groupby and aggregation, merges, dates and resampling, matplotlib and seaborn charts, and notebooks as reports.

Every edition is public at its own address.

The shared chapters

What analysts do (7 concepts)

What a data analyst does, a good business question, metrics and KPIs, types of analysis, the analysis workflow, tidy data, and personal data and privacy. The job before the tools.

Spreadsheets properly (9 concepts)

Relative and absolute references, structured tables, SUMIFS and COUNTIFS, XLOOKUP, text and date cleaning functions, FILTER, UNIQUE and SORT, pivot tables, data validation, and the raw, working and output tab layout. Six of these come with formulas.

SQL for analysis (10 concepts)

SELECT, WHERE and ORDER BY, GROUP BY and HAVING, joins, join fan-out, NULL handling, CASE WHEN, CTEs and subqueries, window functions for ranking, running totals and LAG, and grouping by date periods. Nearly every concept has a query.

Statistics you need (10 concepts)

Mean, median and mode, standard deviation and percentiles, distributions and skew, outliers, samples and sampling bias, correlation versus causation, Simpson's paradox, the A/B test, the confidence interval, and p-values and significance.

Cleaning and modelling data (10 concepts)

Data profiling, missing values, duplicates and keys, data types and parsing, standardising categories, the grain of a table, facts, dimensions and the star schema, ratios (sum first, then divide), reconciliation, and the cleaning log.

Charts, dashboards and storytelling (11 concepts)

Choosing a chart, bars and lines first, removing clutter, colour with purpose, accessible charts, misleading charts, dashboard layout, titles that state the point, the one-sentence finding, situation, complication and resolution, and presenting uncertainty and caveats.

Portfolio and first analyst role (9 concepts)

Public datasets, a question-driven portfolio project, the project write-up, analyst resume bullets, analyst hiring rounds, take-home tasks, adjacent and internal routes, freelance data work, and certificates, honestly.

Rules of thumb and glossary

Each edition has nine rules of thumb, eight shared and one for your tool:

• Write the question and the metric definitions before you open any tool.
• Check row counts before and after every join; a total that jumps is usually fan-out.
• For ratios, sum the parts first and divide once; never average percentages.
• Plot a histogram before you quote an average, and use the median when data is skewed.
• Tie your headline numbers to an independent source before anyone else sees them.
• Every chart title should state the takeaway, not the axes.
• Use grey for context and one colour for the point you want noticed.
• Never publish real company, client or personal data in a portfolio, public dashboard or AI chat.

The tool rules are practical: Power BI Desktop is Windows only, anything on Tableau Public can be downloaded by anyone, and a notebook should be restarted and run from the top before you share it. The A to Z glossary adds 56 to 59 smaller terms, depending on the edition.

How to use it with the roadmap

• After each stage: read the matching chapter and answer each concept aloud. If you cannot define a metric or explain a join's fan-out without notes, go back.
• Before an analyst technical round or take-home task: revise the SQL and cleaning chapters, and the portfolio chapter has concepts on hiring rounds and take-home tasks themselves.
• While building a dashboard: keep the charts chapter open and check every title against the "state the point" rule.
• Before presenting: reread the one-sentence finding and presenting uncertainty concepts.

How it works on Sefism

The public page at /cheatsheets/data-analytics/ shows every chapter's title and introduction, with the first chapters open in full. Everything on Sefism is free to learn from. Keeping a cheat sheet uses credits; new members get starting credits, and anyone learning seriously who runs out can ask for more. Keeping the sheet opens every chapter in the member area.

Following the Data analytics roadmap also unlocks this sheet, and your first roadmap is free. Roadmap topics link directly to the concepts here that cover them.

Related on Sefism

• The Data analytics roadmap this sheet accompanies.
• The SQL and databases cheat sheet if you want more depth on queries and design.
• DAX Guide for the Power BI edition, and Tableau Public for the Tableau edition.
• Online store sales and customer retention dashboard, a project that uses every chapter.
• How do I become a business analyst in tech?

Frequently asked questions

Does the sheet work for Google Sheets as well as Excel? The spreadsheets chapter is written for Excel or Google Sheets, matching the roadmap.

I have not chosen a tool yet. Which edition should I read? Read the shared chapters in any edition first; they are the same everywhere. Then compare the three tool chapters before committing on the roadmap.

Is there enough SQL here to pass an analyst SQL test? The SQL chapter covers the core analysis patterns: joins and fan-out, NULLs, CTEs, window functions and running totals. Practise them by hand, as the roadmap advises, rather than only reading them.