What an analyst actually does

Strip away the job-ad language and a data analyst does one thing: they take a question someone cares about — why are sales falling in Multan? which course brings the most students? did the discount campaign pay for itself? — and produce an answer the asker can act on, with evidence behind it. Not a feeling, not a guess dressed in a chart: an answer, with its basis shown.

Notice what that definition includes and excludes. It includes a lot of unglamorous work — finding the data, discovering it is messy, cleaning it, checking it — which is honestly half the job, and lesson 3 treats it with the respect it deserves. It excludes what beginners fear: advanced mathematics. Working analysts use counting, sums, averages, percentages and comparisons for the great majority of their output; the statistics that go beyond that are learnable when needed and are covered properly in the paid course. The scarce skill is not calculus. It is asking the right question of the right data and refusing to fool yourself or your boss with a bad comparison.

Where the work is: banks and telecoms employ floors of analysts in Karachi and Lahore — reporting, credit, churn, branch performance. Every large retailer, courier, school system and hospital group produces management reports monthly, and someone builds them. In our employer research, Data Analytics matched 11 of the 100 researched employers with 80% depth — among the strongest genuine-demand signals in our whole catalogue — and MIS/reporting roles are among the most commonly advertised office jobs in the country. There is also a humbler, nearer version of the job: the person in any SME who can answer 'which product actually makes us money?' from the sales register. That person is an analyst, whatever their card says, and several lessons here use exactly that setting.

The course's plan: the analyst's question discipline (lesson 2), cleaning (lesson 3), the core techniques in Excel (lesson 4), charts that tell the truth (lesson 5), and delivering findings (lesson 6). One dataset runs through it: a small academy's enrolment and fee register — 400 rows, the kind of file every real institution in Pakistan has somewhere.