What analysts actually spend their time on

Most "data analysis" jobs are not machine learning. They are: someone sends you a messy export, asks a vague question, and needs an answer they can act on by Friday.

The job, honestly described

You will spend your time roughly like this:

Notice that cleaning is the biggest block, and explaining is as big as analysing. Courses that skip both are selling you the fun part only.

What "messy" looks like in real life

A sales export from a Nigerian POS system will typically arrive with:

• Dates stored as text, so they sort as 1 Jan, 10 Jan, 2 Jan
• Naira amounts with ₦ and commas, so Excel treats them as words and refuses to sum them
• Customer names typed four different ways: Chidera O., chidera okeke, CHIDERA OKEKE, Chidera Okeke with a double space
• A "Total" row sitting in the middle of the data, which will quietly corrupt every average you calculate

None of that is exotic. That is a normal Tuesday.

The one habit that separates analysts from spreadsheet users

Never edit your raw data. Keep the original export untouched on one sheet, do your cleaning on another, and analyse on a third.

The reason is not tidiness. It is that in three weeks someone will ask "where did this number come from?", and if you edited the original you cannot answer. An answer you cannot defend is worth nothing.

What you'll be able to do

By the end you'll take a raw export, clean it, answer a business question with it, build a dashboard someone can read without you in the room, and write the summary that makes them act.

Try it yourself

Find any spreadsheet you have — bank statement, church records, sales log. Spot three things that would break an analysis:

1. Is any number stored as text?
2. Is the same thing written more than one way?
3. Is there a total or subtotal sitting inside the data?

If you found all three, you already understand why cleaning is half the job.