IHETC Academy
Becoming a data analyst
They spend most of their time cleaning and checking, not modelling. The skill that sets a good analyst apart is not technical: it is the ability to tell someone expecting a conclusion that the data does not support one.
The job as it is actually done
They spend most of their time cleaning and checking, not modelling. The skill that sets a good analyst apart is not technical: it is the ability to tell someone expecting a conclusion that the data does not support one.
- steps
- 4
- of actual work
- 115 h
- fields covered
- 3
Work delivered, measured as such.
What the market says
Demand is real but the door is narrow: organisations first hire domain people who can analyse, rather than analysts looking for a domain. That is why this path starts with data and not with tools.
The sequence
Each step builds on the previous one. Following that order is what keeps the path short: each notion arrives when it is needed, and it sticks.
- 01
Clean data
Collection, cleaning, consistency checks. Eighty per cent of the real work, and the step everyone skips. — about 30 hours of work.
- 02
Analysis and honest reporting
Useful statistics, visualisation that does not lie, stating uncertainty. — about 35 hours of work.
- 03
The business behind the numbers
Understanding the activity being analysed — finance, sales, operations. A context-free analysis serves nobody. — about 25 hours of work.
- 04
Automating without breaking
Making a process reproducible and monitored, knowing where automation becomes risky. — about 25 hours of work.
The fields feeding this path
Each step draws on one field of the base. The count shown is that of the whole field: it states the depth available to go beyond the path itself, once you hold the role.
- AI & data
- 67 (51 %)
- Finance
- 47 (36 %)
- Development
- 17 (13 %)
By the end, you will be able to
- Take a doubtful dataset and establish its reliability
- Produce an analysis that answers the question asked, and say when it cannot
- Build a report readable by a non-technical decision-maker
- Automate a recurring process and monitor its drift
Questions about this path
Do I need to code before starting?
No. The first two steps can be tackled with spreadsheets and no-code tools, which already creates value. Programming becomes useful at the fourth step, and the path introduces it there rather than as a discouraging prerequisite.
Do I need a computer, or is a phone enough?
A computer is needed from the second step onwards. You can read the lessons and follow your progress on a phone, but you do not clean a dataset or build a dashboard on a six-inch screen. We say so before enrolment rather than at the point where the exercise becomes impractical.