By Metridata Editorial Team · Updated October 2026 · 6 min read · Data Analysis
Search for how to become a data analyst and a lot of the advice says to start with Python. The job adverts tell a different story. Most of them ask for Excel, SQL and a dashboard tool long before they ask for code.
What employers actually ask for
An analysis of 328 data analytics job descriptions by AccioJob found Excel in 81% of them, SQL in 60%, Power BI in 43% and Python in 41%. A separate look at 50 Kenyan data analytics adverts, shared on LinkedIn, put Power BI at 61%, SQL at 59% and Python at 53%.
The percentages differ because the samples differ. The pattern does not. The same four tools lead both lists, and the two Kenyan and global samples agree that Python is useful but not usually the first thing a hiring manager checks.
The Kenyan sample also found something that is easy to overlook. Problem-solving appeared in every advert, and communication and business understanding appeared in 96%. Tools get you through the screening. Those three decide whether you are any good once you start.
A sensible order to learn them
1. Excel. You can see results within days. Learn to clean messy data, use lookups, build pivot tables and draw a chart that makes a point. Many analysts still spend a large share of their week here.
2. SQL. Almost every analyst role involves pulling data from a database, and Excel stops being enough once the data is large or lives in several tables. Learn to filter, join, group and summarise.
3. Power BI. It is free to download, and many Kenyan employers already run Microsoft 365, which makes it the natural dashboard tool. One practical warning: Power BI Desktop runs on Windows only, so Mac users need a plan before they start.
4. Python. Learn it after the first three. It earns its place when you need to automate repetitive work, run statistics or build simple models. Senior adverts ask for more of it: one senior analyst advert posted in Kenya in April 2026 lists advanced SQL, Python or R, and Power BI or Tableau as non-negotiable.
The experience problem
The same LinkedIn review looked at 100 entry-level adverts across Kenyan industries. It found that 62% asked for prior professional experience, and only 38% truly required none. That is frustrating, and the way around it is evidence.
Pick a public dataset, such as Kenya open data or a sales dataset from Kaggle. Ask one specific question of it. Clean the data, answer the question, build a short dashboard and write three paragraphs explaining what you found and what you would do about it. Two or three of those on a public page or in a PDF will tell an employer more than a certificate does.
A realistic four-month plan
- Month 1: Excel, using a real dataset from day one.
- Month 2: SQL. Write queries every day, even short ones.
- Month 3: Power BI, ending in a dashboard you could show an interviewer.
- Month 4: Python basics and one end-to-end project that uses all four tools.
This is a plan, not a promise. Pace depends on how many hours you can give it each week, and nobody can guarantee a job at the end.
If you would rather not do it alone
Our Data Analyst Foundations course covers Excel, SQL and Power BI over eight weeks, and the Career Track adds Python, statistics and a capstone project. Both run online or on campus, and a limited number of scholarship places are available for students. You can join the waitlist for the next cohort. See the programmes and prices, or read our earlier guide on how to learn data analysis fast.
