Skip to content
You Need To Understand This

Answer a real question with numbers

Data

Most people who say they are "not a numbers person" have never been shown the actual job. Analysis is not a chart. It is working out what question is being asked, getting the data into a shape that will not lie to you, cleaning the mess nobody wants to clean, summarising it, checking the number is sane, and saying what it means. This module walks that whole path — a real question, spreadsheets, cleaning, PivotTables, numeracy, SQL, and charts — in the order you actually meet them.

7 lessons1 labs2 hr 18 min

Easier if you have done: Digital Foundation

Lessons

  1. 01

    What question are you actually answering

    Turn a vague request like "look into sales" into a question that has an answer, before you open a single file.

    You keep: Can convert a vague data request into a specific answerable question tied to a decision.

    18 min
  2. 02

    Spreadsheets that do not lie to you

    Structure a sheet so analysis is possible, then use the six formulas that do most real work — including joining two files together.

    You keep: Can lay out data correctly and write SUMIFS, IF, IFERROR and a lookup to answer a business question across two files.

    22 min
  3. 03

    Cleaning the data nobody wants to clean

    Find and fix duplicates, inconsistent spellings, text dates, stray spaces and ambiguous blanks — the work that decides whether your analysis is true.

    You keep: Can take a messy export and make it trustworthy, and can say what was changed and why.

    20 min
  4. 04

    PivotTables, and actually summarising

    Answer "which region, which month, which product" questions in a minute instead of an afternoon — and know when a formula is the better tool.

    You keep: Can build a PivotTable to answer a grouped business question, and can tell when a formula is the right choice instead.

    20 min
  5. 05

    Sanity checking a number

    Work out whether a number is plausible before you repeat it — order of magnitude, the base of a percentage, when the average lies, and why correlation is not cause.

    You keep: Can tell whether a number is plausible, and can spot the four ways numbers are made to mislead.

    18 min
  6. 06

    Asking a database questions

    Write SQL that answers real business questions — filtering, grouping, joining two tables, and handling the NULLs that quietly break your counts.

    You keep: Can write a SELECT with WHERE, GROUP BY and a JOIN to answer a business question, and knows what NULL does to the result.

    22 min
  7. 07

    Charts that make an argument

    Choose the right chart for the claim you are making, label it so it reads alone, and write the summary a manager will actually use.

    You keep: Can pick a chart that fits the claim, label it honestly, and state the finding in one sentence.

    18 min

Labs in this module

All labs